Anthony Barbier | 6ff3b19 | 2017-09-04 18:44:23 +0100 | [diff] [blame] | 1 | /* |
Pablo Tello | d2d5f75 | 2018-06-05 13:37:36 +0100 | [diff] [blame] | 2 | * Copyright (c) 2016-2018 ARM Limited. |
Anthony Barbier | 6ff3b19 | 2017-09-04 18:44:23 +0100 | [diff] [blame] | 3 | * |
| 4 | * SPDX-License-Identifier: MIT |
| 5 | * |
| 6 | * Permission is hereby granted, free of charge, to any person obtaining a copy |
| 7 | * of this software and associated documentation files (the "Software"), to |
| 8 | * deal in the Software without restriction, including without limitation the |
| 9 | * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or |
| 10 | * sell copies of the Software, and to permit persons to whom the Software is |
| 11 | * furnished to do so, subject to the following conditions: |
| 12 | * |
| 13 | * The above copyright notice and this permission notice shall be included in all |
| 14 | * copies or substantial portions of the Software. |
| 15 | * |
| 16 | * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR |
| 17 | * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, |
| 18 | * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE |
| 19 | * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER |
| 20 | * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, |
| 21 | * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE |
| 22 | * SOFTWARE. |
| 23 | */ |
| 24 | #include "helpers.h" |
| 25 | |
| 26 | /** Convert an RGB888 image to RGBX8888 |
| 27 | * |
| 28 | * Global Workgroup Size [ DIV_CEIL(width, 16), height ] |
| 29 | * No offset. |
| 30 | * |
| 31 | * @param[in] input_ptr Pointer to the source image. Supported Format: U8 |
| 32 | * @param[in] input_stride_x Stride of the source image in X dimension (in bytes) |
| 33 | * @param[in] input_step_x input_stride_x * number of elements along X processed per workitem(in bytes) |
| 34 | * @param[in] input_stride_y Stride of the source image in Y dimension (in bytes) |
| 35 | * @param[in] input_step_y input_stride_y * number of elements along Y processed per workitem(in bytes) |
| 36 | * @param[in] input_offset_first_element_in_bytes The offset of the first element in the source image |
| 37 | * @param[out] output_ptr Pointer to the destination image. Supported Format: U8 |
| 38 | * @param[in] output_stride_x Stride of the destination image in X dimension (in bytes) |
| 39 | * @param[in] output_step_x output_stride_x * number of elements along X processed per workitem(in bytes) |
| 40 | * @param[in] output_stride_y Stride of the destination image in Y dimension (in bytes) |
| 41 | * @param[in] output_step_y output_stride_y * number of elements along Y processed per workitem(in bytes) |
| 42 | * @param[in] output_offset_first_element_in_bytes The offset of the first element in the destination image |
| 43 | */ |
| 44 | __kernel void RGB888_to_RGBA8888_bt709( |
| 45 | IMAGE_DECLARATION(input), |
| 46 | IMAGE_DECLARATION(output)) |
| 47 | { |
| 48 | Image in = CONVERT_TO_IMAGE_STRUCT(input); |
| 49 | Image out = CONVERT_TO_IMAGE_STRUCT(output); |
| 50 | |
| 51 | // handle 16 pixels every time |
| 52 | uchar16 rgb_0 = vload16(0, in.ptr); |
| 53 | uchar16 rgb_1 = vload16(0, in.ptr + 16); |
| 54 | uchar16 rgb_2 = vload16(0, in.ptr + 32); |
| 55 | |
| 56 | uchar16 rgba_0 = (uchar16)(rgb_0.s012, 255, rgb_0.s345, 255, rgb_0.s678, 255, rgb_0.s9ab, 255); |
Gian Marco Iodice | 040bffe | 2017-09-29 09:50:28 +0100 | [diff] [blame] | 57 | uchar16 rgba_1 = (uchar16)(rgb_0.scde, 255, rgb_0.sf, rgb_1.s01, 255, rgb_1.s234, 255, rgb_1.s567, 255); |
Anthony Barbier | 6ff3b19 | 2017-09-04 18:44:23 +0100 | [diff] [blame] | 58 | uchar16 rgba_2 = (uchar16)(rgb_1.s89a, 255, rgb_1.sbcd, 255, rgb_1.sef, rgb_2.s0, 255, rgb_2.s123, 255); |
| 59 | uchar16 rgba_3 = (uchar16)(rgb_2.s456, 255, rgb_2.s789, 255, rgb_2.sabc, 255, rgb_2.sdef, 255); |
| 60 | |
| 61 | vstore16(rgba_0, 0, out.ptr); |
| 62 | vstore16(rgba_1, 0, out.ptr + 16); |
| 63 | vstore16(rgba_2, 0, out.ptr + 32); |
| 64 | vstore16(rgba_3, 0, out.ptr + 48); |
| 65 | } |
| 66 | |
Manuel Bottini | acaf21d | 2018-09-26 17:38:19 +0100 | [diff] [blame] | 67 | /** Convert an RGB888 image to U8 |
| 68 | * |
| 69 | * Global Workgroup Size [ DIV_CEIL(width, 16), height ] |
| 70 | * No offset. |
| 71 | * |
| 72 | * @param[in] input_ptr Pointer to the source image. Supported Format: RGB888 |
| 73 | * @param[in] input_stride_x Stride of the source image in X dimension (in bytes) |
| 74 | * @param[in] input_step_x input_stride_x * number of elements along X processed per workitem(in bytes) |
| 75 | * @param[in] input_stride_y Stride of the source image in Y dimension (in bytes) |
| 76 | * @param[in] input_step_y input_stride_y * number of elements along Y processed per workitem(in bytes) |
| 77 | * @param[in] input_offset_first_element_in_bytes The offset of the first element in the source image |
| 78 | * @param[out] output_ptr Pointer to the destination image. Supported Format: U8 |
| 79 | * @param[in] output_stride_x Stride of the destination image in X dimension (in bytes) |
| 80 | * @param[in] output_step_x output_stride_x * number of elements along X processed per workitem(in bytes) |
| 81 | * @param[in] output_stride_y Stride of the destination image in Y dimension (in bytes) |
| 82 | * @param[in] output_step_y output_stride_y * number of elements along Y processed per workitem(in bytes) |
| 83 | * @param[in] output_offset_first_element_in_bytes The offset of the first element in the destination image |
| 84 | */ |
| 85 | __kernel void RGB888_to_U8_bt709( |
| 86 | IMAGE_DECLARATION(input), |
| 87 | IMAGE_DECLARATION(output)) |
| 88 | { |
| 89 | Image in = CONVERT_TO_IMAGE_STRUCT(input); |
| 90 | Image out = CONVERT_TO_IMAGE_STRUCT(output); |
| 91 | |
| 92 | // handle 16 pixels every time |
| 93 | const uchar16 rgb_0 = vload16(0, in.ptr); |
| 94 | const uchar16 rgb_1 = vload16(0, in.ptr + 16); |
| 95 | const uchar16 rgb_2 = vload16(0, in.ptr + 32); |
| 96 | |
| 97 | //Resequence values from a sequence of 16 RGB values to sequence of 16 R, 16 G, 16 B values |
| 98 | const uchar16 rgb_r = (uchar16)(rgb_0.s0369, rgb_0.scf, rgb_1.s258b, rgb_1.se, rgb_2.s147a, rgb_2.sd); |
| 99 | const uchar16 rgb_g = (uchar16)(rgb_0.s147a, rgb_0.sd, rgb_1.s0369, rgb_1.scf, rgb_2.s258b, rgb_2.se); |
| 100 | const uchar16 rgb_b = (uchar16)(rgb_0.s258b, rgb_0.se, rgb_1.s147a, rgb_1.sd, rgb_2.s0369, rgb_2.scf); |
| 101 | |
| 102 | const float16 rgb2u8_red_coef_bt709 = 0.2126f; |
| 103 | const float16 rgb2u8_green_coef_bt709 = 0.7152f; |
| 104 | const float16 rgb2u8_blue_coef_bt709 = 0.0722f; |
| 105 | |
| 106 | //Computation of 16 greyscale values in float |
| 107 | const float16 greyscale_f_0 = rgb2u8_red_coef_bt709 * convert_float16(rgb_r) + rgb2u8_green_coef_bt709 * convert_float16(rgb_g) + rgb2u8_blue_coef_bt709 * convert_float16(rgb_b); |
| 108 | |
| 109 | //Convert it to 16 grayscale uchar values |
| 110 | const uchar16 greyscale_u8_0 = convert_uchar16_sat_rtz(greyscale_f_0); |
| 111 | |
| 112 | vstore16(greyscale_u8_0, 0, out.ptr); |
| 113 | } |
| 114 | |
Anthony Barbier | 6ff3b19 | 2017-09-04 18:44:23 +0100 | [diff] [blame] | 115 | /** Convert an RGB888 image to RGBX8888 |
| 116 | * |
| 117 | * Global Workgroup Size [ DIV_CEIL(width, 16), height ] |
| 118 | * No offset. |
| 119 | * |
| 120 | * @param[in] input_ptr Pointer to the source image. Supported Format: U8 |
| 121 | * @param[in] input_stride_x Stride of the source image in X dimension (in bytes) |
| 122 | * @param[in] input_step_x input_stride_x * number of elements along X processed per workitem(in bytes) |
| 123 | * @param[in] input_stride_y Stride of the source image in Y dimension (in bytes) |
| 124 | * @param[in] input_step_y input_stride_y * number of elements along Y processed per workitem(in bytes) |
| 125 | * @param[in] input_offset_first_element_in_bytes The offset of the first element in the source image |
| 126 | * @param[out] output_ptr Pointer to the destination image. Supported Format: U8 |
| 127 | * @param[in] output_stride_x Stride of the destination image in X dimension (in bytes) |
| 128 | * @param[in] output_step_x output_stride_x * number of elements along X processed per workitem(in bytes) |
| 129 | * @param[in] output_stride_y Stride of the destination image in Y dimension (in bytes) |
| 130 | * @param[in] output_step_y output_stride_y * number of elements along Y processed per workitem(in bytes) |
| 131 | * @param[in] output_offset_first_element_in_bytes The offset of the first element in the destination image |
| 132 | */ |
| 133 | __kernel void RGBA8888_to_RGB888_bt709( |
| 134 | IMAGE_DECLARATION(input), |
| 135 | IMAGE_DECLARATION(output)) |
| 136 | { |
| 137 | Image in = CONVERT_TO_IMAGE_STRUCT(input); |
| 138 | Image out = CONVERT_TO_IMAGE_STRUCT(output); |
| 139 | // handle 16 pixels every time |
| 140 | uchar16 rgba_0 = vload16(0, in.ptr); |
| 141 | uchar16 rgba_1 = vload16(0, in.ptr + 16); |
| 142 | uchar16 rgba_2 = vload16(0, in.ptr + 32); |
| 143 | uchar16 rgba_3 = vload16(0, in.ptr + 48); |
| 144 | |
| 145 | uchar16 rgb_0 = (uchar16)(rgba_0.s01245689, rgba_0.sacde, rgba_1.s0124); |
| 146 | uchar16 rgb_1 = (uchar16)(rgba_1.s5689acde, rgba_2.s01245689); |
| 147 | uchar16 rgb_2 = (uchar16)(rgba_2.sacde, rgba_3.s01245689, rgba_3.sacde); |
| 148 | |
| 149 | vstore16(rgb_0, 0, out.ptr); |
| 150 | vstore16(rgb_1, 0, out.ptr + 16); |
| 151 | vstore16(rgb_2, 0, out.ptr + 32); |
| 152 | } |
| 153 | |
| 154 | /** Convert a UYVY422 image to RGB888 using BT709 color space |
| 155 | * |
| 156 | * Global Workgroup Size [ DIV_CEIL(width, 8), height ] |
| 157 | * No offset. |
| 158 | * |
| 159 | * @param[in] input_ptr Pointer to the source image. Supported Format: U8 |
| 160 | * @param[in] input_stride_x Stride of the source image in X dimension (in bytes) |
| 161 | * @param[in] input_step_x input_stride_x * number of elements along X processed per workitem(in bytes) |
| 162 | * @param[in] input_stride_y Stride of the source image in Y dimension (in bytes) |
| 163 | * @param[in] input_step_y input_stride_y * number of elements along Y processed per workitem(in bytes) |
| 164 | * @param[in] input_offset_first_element_in_bytes The offset of the first element in the source image |
| 165 | * @param[out] output_ptr Pointer to the destination image. Supported Format: U8 |
| 166 | * @param[in] output_stride_x Stride of the destination image in X dimension (in bytes) |
| 167 | * @param[in] output_step_x output_stride_x * number of elements along X processed per workitem(in bytes) |
| 168 | * @param[in] output_stride_y Stride of the destination image in Y dimension (in bytes) |
| 169 | * @param[in] output_step_y output_stride_y * number of elements along Y processed per workitem(in bytes) |
| 170 | * @param[in] output_offset_first_element_in_bytes The offset of the first element in the destination image |
| 171 | */ |
| 172 | __kernel void UYVY422_to_RGB888_bt709( |
| 173 | IMAGE_DECLARATION(input), |
| 174 | IMAGE_DECLARATION(output)) |
| 175 | { |
| 176 | Image in = CONVERT_TO_IMAGE_STRUCT(input); |
| 177 | Image out = CONVERT_TO_IMAGE_STRUCT(output); |
| 178 | |
| 179 | // handle 8 pixels every time |
| 180 | uchar16 uyvy = vload16(0, in.ptr); |
| 181 | |
| 182 | uchar8 luma = (uchar8)(uyvy.s1, uyvy.s3, uyvy.s5, uyvy.s7, uyvy.s9, uyvy.sb, uyvy.sd, uyvy.sf); |
| 183 | char8 cb = (char8)(uyvy.s0, uyvy.s0, uyvy.s4, uyvy.s4, uyvy.s8, uyvy.s8, uyvy.sc, uyvy.sc) - (char8)(128); |
| 184 | char8 cr = (char8)(uyvy.s2, uyvy.s2, uyvy.s6, uyvy.s6, uyvy.sa, uyvy.sa, uyvy.se, uyvy.se) - (char8)(128); |
| 185 | |
Pablo Tello | d2d5f75 | 2018-06-05 13:37:36 +0100 | [diff] [blame] | 186 | float8 red_coef_bt709 = (float8)(1.5748f); |
| 187 | float8 green_coef_bt709 = (float8)(-0.1873f); |
| 188 | float8 green_coef2_bt709 = (float8)(-0.4681f); |
| 189 | float8 blue_coef_bt709 = (float8)(1.8556f); |
| 190 | float8 lumav = convert_float8(luma); |
| 191 | |
| 192 | float8 f_r = red_coef_bt709 * convert_float8(cr); |
| 193 | float8 f_g = green_coef_bt709 * convert_float8(cb) + green_coef2_bt709 * convert_float8(cr); |
| 194 | float8 f_b = blue_coef_bt709 * convert_float8(cb); |
| 195 | |
| 196 | f_r += lumav; |
| 197 | f_g += lumav; |
| 198 | f_b += lumav; |
Anthony Barbier | 6ff3b19 | 2017-09-04 18:44:23 +0100 | [diff] [blame] | 199 | |
Pablo Tello | 96fc1d6 | 2018-07-17 17:10:59 +0100 | [diff] [blame] | 200 | uchar8 r_0 = convert_uchar8_sat_rtz(f_r); |
| 201 | uchar8 g_0 = convert_uchar8_sat_rtz(f_g); |
| 202 | uchar8 b_0 = convert_uchar8_sat_rtz(f_b); |
Anthony Barbier | 6ff3b19 | 2017-09-04 18:44:23 +0100 | [diff] [blame] | 203 | |
| 204 | uchar16 rgb_0 = (uchar16)(r_0.s0, g_0.s0, b_0.s0, r_0.s1, g_0.s1, b_0.s1, r_0.s2, g_0.s2, b_0.s2, |
| 205 | r_0.s3, g_0.s3, b_0.s3, r_0.s4, g_0.s4, b_0.s4, r_0.s5); |
| 206 | uchar8 rgb_1 = (uchar8)(g_0.s5, b_0.s5, r_0.s6, g_0.s6, b_0.s6, r_0.s7, g_0.s7, b_0.s7); |
| 207 | |
| 208 | vstore16(rgb_0, 0, out.ptr); |
| 209 | vstore8(rgb_1, 0, out.ptr + 16); |
| 210 | } |
| 211 | |
| 212 | /** Convert a UYVY422 image to RGBX8888 using BT709 color space |
| 213 | * |
| 214 | * Global Workgroup Size [ DIV_CEIL(width, 8), height ] |
| 215 | * No offset. |
| 216 | * |
| 217 | * @param[in] input_ptr Pointer to the source image. Supported Format: U8 |
| 218 | * @param[in] input_stride_x Stride of the source image in X dimension (in bytes) |
| 219 | * @param[in] input_step_x input_stride_x * number of elements along X processed per workitem(in bytes) |
| 220 | * @param[in] input_stride_y Stride of the source image in Y dimension (in bytes) |
| 221 | * @param[in] input_step_y input_stride_y * number of elements along Y processed per workitem(in bytes) |
| 222 | * @param[in] input_offset_first_element_in_bytes The offset of the first element in the source image |
| 223 | * @param[out] output_ptr Pointer to the destination image. Supported Format: U8 |
| 224 | * @param[in] output_stride_x Stride of the destination image in X dimension (in bytes) |
| 225 | * @param[in] output_step_x output_stride_x * number of elements along X processed per workitem(in bytes) |
| 226 | * @param[in] output_stride_y Stride of the destination image in Y dimension (in bytes) |
| 227 | * @param[in] output_step_y output_stride_y * number of elements along Y processed per workitem(in bytes) |
| 228 | * @param[in] output_offset_first_element_in_bytes The offset of the first element in the destination image |
| 229 | */ |
| 230 | __kernel void UYVY422_to_RGBA8888_bt709( |
| 231 | IMAGE_DECLARATION(input), |
| 232 | IMAGE_DECLARATION(output)) |
| 233 | { |
| 234 | Image in = CONVERT_TO_IMAGE_STRUCT(input); |
| 235 | Image out = CONVERT_TO_IMAGE_STRUCT(output); |
| 236 | |
| 237 | // handle 8 pixels every time |
| 238 | uchar16 uyvy = vload16(0, in.ptr); |
| 239 | |
| 240 | uchar8 luma = (uchar8)(uyvy.s1, uyvy.s3, uyvy.s5, uyvy.s7, uyvy.s9, uyvy.sb, uyvy.sd, uyvy.sf); |
| 241 | char8 cb = (char8)(uyvy.s0, uyvy.s0, uyvy.s4, uyvy.s4, uyvy.s8, uyvy.s8, uyvy.sc, uyvy.sc) - (char8)(128); |
| 242 | char8 cr = (char8)(uyvy.s2, uyvy.s2, uyvy.s6, uyvy.s6, uyvy.sa, uyvy.sa, uyvy.se, uyvy.se) - (char8)(128); |
| 243 | |
Pablo Tello | d2d5f75 | 2018-06-05 13:37:36 +0100 | [diff] [blame] | 244 | float8 red_coef_bt709 = (float8)(1.5748f); |
| 245 | float8 green_coef_bt709 = (float8)(-0.1873f); |
| 246 | float8 green_coef2_bt709 = (float8)(-0.4681f); |
| 247 | float8 blue_coef_bt709 = (float8)(1.8556f); |
| 248 | float8 lumav = convert_float8(luma); |
| 249 | |
| 250 | float8 f_r = red_coef_bt709 * convert_float8(cr); |
| 251 | float8 f_g = green_coef_bt709 * convert_float8(cb) + green_coef2_bt709 * convert_float8(cr); |
| 252 | float8 f_b = blue_coef_bt709 * convert_float8(cb); |
| 253 | |
| 254 | f_r += lumav; |
| 255 | f_g += lumav; |
| 256 | f_b += lumav; |
Anthony Barbier | 6ff3b19 | 2017-09-04 18:44:23 +0100 | [diff] [blame] | 257 | |
Pablo Tello | 96fc1d6 | 2018-07-17 17:10:59 +0100 | [diff] [blame] | 258 | uchar8 r_0 = convert_uchar8_sat_rtz(f_r); |
| 259 | uchar8 g_0 = convert_uchar8_sat_rtz(f_g); |
| 260 | uchar8 b_0 = convert_uchar8_sat_rtz(f_b); |
Anthony Barbier | 6ff3b19 | 2017-09-04 18:44:23 +0100 | [diff] [blame] | 261 | |
| 262 | uchar16 rgba_0 = (uchar16)(r_0.s0, g_0.s0, b_0.s0, 255, r_0.s1, g_0.s1, b_0.s1, 255, |
| 263 | r_0.s2, g_0.s2, b_0.s2, 255, r_0.s3, g_0.s3, b_0.s3, 255); |
| 264 | uchar16 rgba_1 = (uchar16)(r_0.s4, g_0.s4, b_0.s4, 255, r_0.s5, g_0.s5, b_0.s5, 255, |
| 265 | r_0.s6, g_0.s6, b_0.s6, 255, r_0.s7, g_0.s7, b_0.s7, 255); |
| 266 | |
| 267 | vstore16(rgba_0, 0, out.ptr); |
| 268 | vstore16(rgba_1, 0, out.ptr + 16); |
| 269 | } |
| 270 | |
| 271 | /** Convert a YUYV422 image to RGB888 using BT709 color space |
| 272 | * |
| 273 | * Global Workgroup Size [ DIV_CEIL(width, 8), height ] |
| 274 | * No offset. |
| 275 | * |
| 276 | * @param[in] input_ptr Pointer to the source image. Supported Format: U8 |
| 277 | * @param[in] input_stride_x Stride of the source image in X dimension (in bytes) |
| 278 | * @param[in] input_step_x input_stride_x * number of elements along X processed per workitem(in bytes) |
| 279 | * @param[in] input_stride_y Stride of the source image in Y dimension (in bytes) |
| 280 | * @param[in] input_step_y input_stride_y * number of elements along Y processed per workitem(in bytes) |
| 281 | * @param[in] input_offset_first_element_in_bytes The offset of the first element in the source image |
| 282 | * @param[out] output_ptr Pointer to the destination image. Supported Format: U8 |
| 283 | * @param[in] output_stride_x Stride of the destination image in X dimension (in bytes) |
| 284 | * @param[in] output_step_x output_stride_x * number of elements along X processed per workitem(in bytes) |
| 285 | * @param[in] output_stride_y Stride of the destination image in Y dimension (in bytes) |
| 286 | * @param[in] output_step_y output_stride_y * number of elements along Y processed per workitem(in bytes) |
| 287 | * @param[in] output_offset_first_element_in_bytes The offset of the first element in the destination image |
| 288 | */ |
| 289 | __kernel void YUYV422_to_RGB888_bt709( |
| 290 | IMAGE_DECLARATION(input), |
| 291 | IMAGE_DECLARATION(output)) |
| 292 | { |
| 293 | Image in = CONVERT_TO_IMAGE_STRUCT(input); |
| 294 | Image out = CONVERT_TO_IMAGE_STRUCT(output); |
| 295 | |
| 296 | // handle 8 pixels every time |
| 297 | uchar16 uyvy = vload16(0, in.ptr); |
| 298 | |
| 299 | uchar8 luma = (uchar8)(uyvy.s0, uyvy.s2, uyvy.s4, uyvy.s6, uyvy.s8, uyvy.sa, uyvy.sc, uyvy.se); |
| 300 | char8 cb = (char8)(uyvy.s1, uyvy.s1, uyvy.s5, uyvy.s5, uyvy.s9, uyvy.s9, uyvy.sd, uyvy.sd) - (char8)(128); |
| 301 | char8 cr = (char8)(uyvy.s3, uyvy.s3, uyvy.s7, uyvy.s7, uyvy.sb, uyvy.sb, uyvy.sf, uyvy.sf) - (char8)(128); |
| 302 | |
Pablo Tello | d2d5f75 | 2018-06-05 13:37:36 +0100 | [diff] [blame] | 303 | float8 red_coef_bt709 = (float8)(1.5748f); |
| 304 | float8 green_coef_bt709 = (float8)(-0.1873f); |
| 305 | float8 green_coef2_bt709 = (float8)(-0.4681f); |
| 306 | float8 blue_coef_bt709 = (float8)(1.8556f); |
| 307 | float8 lumav = convert_float8(luma); |
| 308 | |
| 309 | float8 f_r = red_coef_bt709 * convert_float8(cr); |
| 310 | float8 f_g = green_coef_bt709 * convert_float8(cb) + green_coef2_bt709 * convert_float8(cr); |
| 311 | float8 f_b = blue_coef_bt709 * convert_float8(cb); |
| 312 | |
| 313 | f_r += lumav; |
| 314 | f_g += lumav; |
| 315 | f_b += lumav; |
Anthony Barbier | 6ff3b19 | 2017-09-04 18:44:23 +0100 | [diff] [blame] | 316 | |
Pablo Tello | 96fc1d6 | 2018-07-17 17:10:59 +0100 | [diff] [blame] | 317 | uchar8 r_0 = convert_uchar8_sat_rtz(f_r); |
| 318 | uchar8 g_0 = convert_uchar8_sat_rtz(f_g); |
| 319 | uchar8 b_0 = convert_uchar8_sat_rtz(f_b); |
Anthony Barbier | 6ff3b19 | 2017-09-04 18:44:23 +0100 | [diff] [blame] | 320 | |
| 321 | uchar16 rgb_0 = (uchar16)(r_0.s0, g_0.s0, b_0.s0, r_0.s1, g_0.s1, b_0.s1, r_0.s2, g_0.s2, b_0.s2, |
| 322 | r_0.s3, g_0.s3, b_0.s3, r_0.s4, g_0.s4, b_0.s4, r_0.s5); |
| 323 | uchar8 rgb_1 = (uchar8)(g_0.s5, b_0.s5, r_0.s6, g_0.s6, b_0.s6, r_0.s7, g_0.s7, b_0.s7); |
| 324 | |
| 325 | vstore16(rgb_0, 0, out.ptr); |
| 326 | vstore8(rgb_1, 0, out.ptr + 16); |
| 327 | } |
| 328 | |
| 329 | /** Convert a YUYV422 image to RGBX8888 using BT709 color space |
| 330 | * |
| 331 | * Global Workgroup Size [ DIV_CEIL(width, 8), height ] |
| 332 | * No offset. |
| 333 | * |
| 334 | * @param[in] input_ptr Pointer to the source image. Supported Format: U8 |
| 335 | * @param[in] input_stride_x Stride of the source image in X dimension (in bytes) |
| 336 | * @param[in] input_step_x input_stride_x * number of elements along X processed per workitem(in bytes) |
| 337 | * @param[in] input_stride_y Stride of the source image in Y dimension (in bytes) |
| 338 | * @param[in] input_step_y input_stride_y * number of elements along Y processed per workitem(in bytes) |
| 339 | * @param[in] input_offset_first_element_in_bytes The offset of the first element in the source image |
| 340 | * @param[out] output_ptr Pointer to the destination image. Supported Format: U8 |
| 341 | * @param[in] output_stride_x Stride of the destination image in X dimension (in bytes) |
| 342 | * @param[in] output_step_x output_stride_x * number of elements along X processed per workitem(in bytes) |
| 343 | * @param[in] output_stride_y Stride of the destination image in Y dimension (in bytes) |
| 344 | * @param[in] output_step_y output_stride_y * number of elements along Y processed per workitem(in bytes) |
| 345 | * @param[in] output_offset_first_element_in_bytes The offset of the first element in the destination image |
| 346 | */ |
| 347 | __kernel void YUYV422_to_RGBA8888_bt709( |
| 348 | IMAGE_DECLARATION(input), |
| 349 | IMAGE_DECLARATION(output)) |
| 350 | { |
| 351 | Image in = CONVERT_TO_IMAGE_STRUCT(input); |
| 352 | Image out = CONVERT_TO_IMAGE_STRUCT(output); |
| 353 | |
| 354 | // handle 8 pixels every time |
| 355 | uchar16 uyvy = vload16(0, in.ptr); |
| 356 | |
| 357 | uchar8 luma = (uchar8)(uyvy.s0, uyvy.s2, uyvy.s4, uyvy.s6, uyvy.s8, uyvy.sa, uyvy.sc, uyvy.se); |
| 358 | char8 cb = (char8)(uyvy.s1, uyvy.s1, uyvy.s5, uyvy.s5, uyvy.s9, uyvy.s9, uyvy.sd, uyvy.sd) - (char8)(128); |
| 359 | char8 cr = (char8)(uyvy.s3, uyvy.s3, uyvy.s7, uyvy.s7, uyvy.sb, uyvy.sb, uyvy.sf, uyvy.sf) - (char8)(128); |
| 360 | |
Pablo Tello | d2d5f75 | 2018-06-05 13:37:36 +0100 | [diff] [blame] | 361 | float8 red_coef_bt709 = (float8)(1.5748f); |
| 362 | float8 green_coef_bt709 = (float8)(-0.1873f); |
| 363 | float8 green_coef2_bt709 = (float8)(-0.4681f); |
| 364 | float8 blue_coef_bt709 = (float8)(1.8556f); |
| 365 | float8 lumav = convert_float8(luma); |
| 366 | |
| 367 | float8 f_r = red_coef_bt709 * convert_float8(cr); |
| 368 | float8 f_g = green_coef_bt709 * convert_float8(cb) + green_coef2_bt709 * convert_float8(cr); |
| 369 | float8 f_b = blue_coef_bt709 * convert_float8(cb); |
| 370 | |
| 371 | f_r += lumav; |
| 372 | f_g += lumav; |
| 373 | f_b += lumav; |
Anthony Barbier | 6ff3b19 | 2017-09-04 18:44:23 +0100 | [diff] [blame] | 374 | |
Pablo Tello | 96fc1d6 | 2018-07-17 17:10:59 +0100 | [diff] [blame] | 375 | uchar8 r_0 = convert_uchar8_sat_rtz(f_r); |
| 376 | uchar8 g_0 = convert_uchar8_sat_rtz(f_g); |
| 377 | uchar8 b_0 = convert_uchar8_sat_rtz(f_b); |
Anthony Barbier | 6ff3b19 | 2017-09-04 18:44:23 +0100 | [diff] [blame] | 378 | |
| 379 | uchar16 rgba_0 = (uchar16)(r_0.s0, g_0.s0, b_0.s0, 255, r_0.s1, g_0.s1, b_0.s1, 255, |
| 380 | r_0.s2, g_0.s2, b_0.s2, 255, r_0.s3, g_0.s3, b_0.s3, 255); |
| 381 | uchar16 rgba_1 = (uchar16)(r_0.s4, g_0.s4, b_0.s4, 255, r_0.s5, g_0.s5, b_0.s5, 255, |
| 382 | r_0.s6, g_0.s6, b_0.s6, 255, r_0.s7, g_0.s7, b_0.s7, 255); |
| 383 | |
| 384 | vstore16(rgba_0, 0, out.ptr); |
| 385 | vstore16(rgba_1, 0, out.ptr + 16); |
| 386 | } |
| 387 | |
| 388 | /** Convert a RGB image to NV12 using BT709 color space |
| 389 | * |
| 390 | * Global Workgroup Size [ DIV_CEIL(width, 4), height ] |
| 391 | * No offset. |
| 392 | * |
| 393 | * @param[in] input_ptr Pointer to the source image. Supported Format: U8 |
| 394 | * @param[in] input_stride_x Stride of the source image in X dimension (in bytes) |
| 395 | * @param[in] input_step_x input_stride_x * number of elements along X processed per workitem(in bytes) |
| 396 | * @param[in] input_stride_y Stride of the source image in Y dimension (in bytes) |
| 397 | * @param[in] input_step_y input_stride_y * number of elements along Y processed per workitem(in bytes) |
| 398 | * @param[in] input_offset_first_element_in_bytes The offset of the first element in the source image |
| 399 | * @param[out] luma_ptr Pointer to the destination luma channel. Supported Format: U8 |
| 400 | * @param[in] luma_stride_x Stride of the destination luma channel in X dimension (in bytes) |
| 401 | * @param[in] luma_step_x luma_stride_x * number of elements along X processed per workitem(in bytes) |
| 402 | * @param[in] luma_stride_y Stride of the destination image luma channel in Y dimension (in bytes) |
| 403 | * @param[in] luma_step_y luma_stride_y * number of elements along Y processed per workitem(in bytes) |
| 404 | * @param[in] luma_offset_first_element_in_bytes The offset of the first element in the destination image luma channel |
| 405 | * @param[out] uv_ptr Pointer to the destination uv channel. Supported Format: U8 |
| 406 | * @param[in] uv_stride_x Stride of the destination uv channel in X dimension (in bytes) |
| 407 | * @param[in] uv_step_x uv_stride_x * number of elements along X processed per workitem(in bytes) |
| 408 | * @param[in] uv_stride_y Stride of the destination image luma channel in Y dimension (in bytes) |
| 409 | * @param[in] uv_step_y uv_stride_y * number of elements along Y processed per workitem(in bytes) |
| 410 | * @param[in] uv_offset_first_element_in_bytes The offset of the first element in the destination image uv channel |
| 411 | * |
| 412 | */ |
| 413 | __kernel void RGB888_to_NV12_bt709( |
| 414 | IMAGE_DECLARATION(input), |
| 415 | IMAGE_DECLARATION(luma), |
| 416 | IMAGE_DECLARATION(uv)) |
| 417 | { |
| 418 | Image in = CONVERT_TO_IMAGE_STRUCT(input); |
| 419 | Image out_y = CONVERT_TO_IMAGE_STRUCT(luma); |
| 420 | Image out_uv = CONVERT_TO_IMAGE_STRUCT(uv); |
| 421 | |
| 422 | // handle 4 pixels every time, two lines, each line for 2 pixels |
| 423 | // Read 2 pixel of the first line |
| 424 | uchar8 rgb_0 = vload8(0, in.ptr); |
| 425 | uchar2 r_0 = (uchar2)(rgb_0.s0, rgb_0.s3); |
| 426 | uchar2 g_0 = (uchar2)(rgb_0.s1, rgb_0.s4); |
| 427 | uchar2 b_0 = (uchar2)(rgb_0.s2, rgb_0.s5); |
| 428 | |
| 429 | float2 f_y = (float2)(0.0000f) + (float2)(0.2126f) * convert_float2(r_0) + (float2)(0.7152f) * convert_float2(g_0) + (float2)(0.0722f) * convert_float2(b_0); |
| 430 | float2 f_u = (float2)(0.0000f) - (float2)(0.1146f) * convert_float2(r_0) - (float2)(0.3854f) * convert_float2(g_0) + (float2)(0.5000f) * convert_float2(b_0); |
| 431 | float2 f_v = (float2)(0.0000f) + (float2)(0.5000f) * convert_float2(r_0) - (float2)(0.4542f) * convert_float2(g_0) - (float2)(0.0458f) * convert_float2(b_0); |
| 432 | |
| 433 | short2 i_y = convert_short2_rtz(f_y); |
| 434 | short2 i_u = convert_short2_rtz(f_u) + (short2)(128); |
| 435 | short2 i_v = convert_short2_rtz(f_v) + (short2)(128); |
| 436 | |
| 437 | uchar2 luma_0 = convert_uchar2(max((short2)(0), min(i_y, (short2)(255)))); |
| 438 | vstore2(luma_0, 0, out_y.ptr); |
| 439 | |
| 440 | uchar2 cb_0 = convert_uchar2(max((short2)(0), min(i_u, (short2)(255)))); |
| 441 | uchar2 cr_0 = convert_uchar2(max((short2)(0), min(i_v, (short2)(255)))); |
| 442 | |
| 443 | // Read 2 pixel of the second line |
| 444 | uchar8 rgb_1 = vload8(0, in.ptr + input_stride_y); |
| 445 | uchar2 r_1 = (uchar2)(rgb_1.s0, rgb_1.s3); |
| 446 | uchar2 g_1 = (uchar2)(rgb_1.s1, rgb_1.s4); |
| 447 | uchar2 b_1 = (uchar2)(rgb_1.s2, rgb_1.s5); |
| 448 | |
| 449 | f_y = (float2)(0.0000f) + (float2)(0.2126f) * convert_float2(r_1) + (float2)(0.7152f) * convert_float2(g_1) + (float2)(0.0722f) * convert_float2(b_1); |
| 450 | f_u = (float2)(0.0000f) - (float2)(0.1146f) * convert_float2(r_1) - (float2)(0.3854f) * convert_float2(g_1) + (float2)(0.5000f) * convert_float2(b_1); |
| 451 | f_v = (float2)(0.0000f) + (float2)(0.5000f) * convert_float2(r_1) - (float2)(0.4542f) * convert_float2(g_1) - (float2)(0.0458f) * convert_float2(b_1); |
| 452 | |
| 453 | i_y = convert_short2_rtz(f_y); |
| 454 | i_u = convert_short2_rtz(f_u) + (short2)(128); |
| 455 | i_v = convert_short2_rtz(f_v) + (short2)(128); |
| 456 | |
| 457 | uchar2 luma_1 = convert_uchar2(max((short2)(0), min(i_y, (short2)(255)))); |
| 458 | vstore2(luma_1, 0, out_y.ptr + luma_stride_y); |
| 459 | |
| 460 | uchar2 cb_1 = convert_uchar2(max((short2)(0), min(i_u, (short2)(255)))); |
| 461 | uchar2 cr_1 = convert_uchar2(max((short2)(0), min(i_v, (short2)(255)))); |
| 462 | uchar2 cbcr = (uchar2)(((cb_0.s0 + cb_0.s1 + cb_1.s0 + cb_1.s1) / 4), |
| 463 | ((cr_0.s0 + cr_0.s1 + cr_1.s0 + cr_1.s1) / 4)); |
| 464 | |
| 465 | vstore2(cbcr, 0, out_uv.ptr); |
| 466 | } |
| 467 | |
| 468 | /* |
| 469 | R'= Y' + 0.0000*U + 1.5748*V |
| 470 | G'= Y' - 0.1873*U - 0.4681*V |
| 471 | B'= Y' + 1.8556*U + 0.0000*V |
| 472 | */ |
| 473 | |
| 474 | /** Convert an NV12 image to RGB888 |
| 475 | * |
| 476 | * Global Workgroup Size [ DIV_CEIL(width, 4), height ] |
| 477 | * No offset. |
| 478 | * |
| 479 | * @param[in] luma_input_ptr Pointer to the source luma channel. Supported Format: U8 |
| 480 | * @param[in] luma_input_stride_x Stride of the luma image in X dimension (in bytes) |
| 481 | * @param[in] luma_input_step_x luma_input_stride_x * number of elements along X processed per workitem(in bytes) |
| 482 | * @param[in] luma_input_stride_y Stride of the source luma channel in Y dimension (in bytes) |
| 483 | * @param[in] luma_input_step_y luma_input_stride_y * number of elements along Y processed per workitem(in bytes) |
| 484 | * @param[in] luma_input_offset_first_element_in_bytes The offset of the first element in the source image |
| 485 | * @param[in] uv_input_ptr Pointer to the source uv channel. Supported Format: U8 |
| 486 | * @param[in] uv_input_stride_x Stride of the source image uv channel in X dimension (in bytes) |
| 487 | * @param[in] uv_input_step_x uv_input_stride_x * number of elements along X processed per workitem(in bytes) |
| 488 | * @param[in] uv_input_stride_y Stride of the source image in Y dimension (in bytes) |
| 489 | * @param[in] uv_input_step_y uv_input_stride_y * number of elements along Y processed per workitem(in bytes) |
| 490 | * @param[in] uv_input_offset_first_element_in_bytes The offset of the first element in the source image |
| 491 | * @param[out] rgb_output_ptr Pointer to the destination image. Supported Format: U8 |
| 492 | * @param[in] rgb_output_stride_x Stride of the destination image in X dimension (in bytes) |
| 493 | * @param[in] rgb_output_step_x rgb_output_stride_x * number of elements along X processed per workitem(in bytes) |
| 494 | * @param[in] rgb_output_stride_y Stride of the destination image in Y dimension (in bytes) |
| 495 | * @param[in] rgb_output_step_y output_stride_y * number of elements along Y processed per workitem(in bytes) |
| 496 | * @param[in] rgb_output_offset_first_element_in_bytes The offset of the first element in the destination image |
| 497 | */ |
| 498 | __kernel void NV12_to_RGB888_bt709( |
| 499 | IMAGE_DECLARATION(luma_input), |
| 500 | IMAGE_DECLARATION(uv_input), |
| 501 | IMAGE_DECLARATION(rgb_output)) |
| 502 | { |
| 503 | Image in_luma = CONVERT_TO_IMAGE_STRUCT(luma_input); |
| 504 | Image in_uv = CONVERT_TO_IMAGE_STRUCT(uv_input); |
| 505 | Image out_rgb = CONVERT_TO_IMAGE_STRUCT(rgb_output); |
| 506 | |
| 507 | // handle 8 pixels every time, two lines, each line for 4 pixels |
| 508 | uchar4 luma_0 = vload4(0, in_luma.ptr); |
| 509 | uchar4 luma_1 = vload4(0, in_luma.ptr + luma_input_stride_y); |
| 510 | uchar4 cbcr = vload4(0, in_uv.ptr); |
| 511 | char4 cb = (char4)(cbcr.s0, cbcr.s0, cbcr.s2, cbcr.s2) - (char4)(128); |
| 512 | char4 cr = (char4)(cbcr.s1, cbcr.s1, cbcr.s3, cbcr.s3) - (char4)(128); |
| 513 | |
| 514 | float4 temp0 = (float4)(0.0000f) + (float4)(0.0000f) * convert_float4(cb) + (float4)(1.5748f) * convert_float4(cr); |
| 515 | float4 temp1 = (float4)(0.0000f) - (float4)(0.1873f) * convert_float4(cb) - (float4)(0.4681f) * convert_float4(cr); |
| 516 | float4 temp2 = (float4)(0.0000f) + (float4)(1.8556f) * convert_float4(cb) + (float4)(0.0000f) * convert_float4(cr); |
| 517 | |
| 518 | float4 f_r = convert_float4(luma_0) + temp0; |
| 519 | float4 f_g = convert_float4(luma_0) + temp1; |
| 520 | float4 f_b = convert_float4(luma_0) + temp2; |
| 521 | |
Pablo Tello | 96fc1d6 | 2018-07-17 17:10:59 +0100 | [diff] [blame] | 522 | uchar4 r_0 = convert_uchar4_sat_rtz(f_r); |
| 523 | uchar4 g_0 = convert_uchar4_sat_rtz(f_g); |
| 524 | uchar4 b_0 = convert_uchar4_sat_rtz(f_b); |
Anthony Barbier | 6ff3b19 | 2017-09-04 18:44:23 +0100 | [diff] [blame] | 525 | |
| 526 | uchar8 rgb_0 = (uchar8)(r_0.s0, g_0.s0, b_0.s0, r_0.s1, g_0.s1, b_0.s1, r_0.s2, g_0.s2); |
| 527 | uchar4 rgb_1 = (uchar4)(b_0.s2, r_0.s3, g_0.s3, b_0.s3); |
| 528 | vstore8(rgb_0, 0, out_rgb.ptr); |
| 529 | vstore4(rgb_1, 0, out_rgb.ptr + 8); |
| 530 | |
| 531 | f_r = convert_float4(luma_1) + temp0; |
| 532 | f_g = convert_float4(luma_1) + temp1; |
| 533 | f_b = convert_float4(luma_1) + temp2; |
| 534 | |
Pablo Tello | 96fc1d6 | 2018-07-17 17:10:59 +0100 | [diff] [blame] | 535 | r_0 = convert_uchar4_sat_rtz(f_r); |
| 536 | g_0 = convert_uchar4_sat_rtz(f_g); |
| 537 | b_0 = convert_uchar4_sat_rtz(f_b); |
Anthony Barbier | 6ff3b19 | 2017-09-04 18:44:23 +0100 | [diff] [blame] | 538 | |
| 539 | rgb_0 = (uchar8)(r_0.s0, g_0.s0, b_0.s0, r_0.s1, g_0.s1, b_0.s1, r_0.s2, g_0.s2); |
| 540 | rgb_1 = (uchar4)(b_0.s2, r_0.s3, g_0.s3, b_0.s3); |
| 541 | vstore8(rgb_0, 0, out_rgb.ptr + rgb_output_stride_y); |
| 542 | vstore4(rgb_1, 0, out_rgb.ptr + rgb_output_stride_y + 8); |
| 543 | } |
| 544 | |
| 545 | /** Convert a RGB image to YUV444 using BT709 color space |
| 546 | * |
| 547 | * Global Workgroup Size [ DIV_CEIL(width, 4), height ] |
| 548 | * No offset. |
| 549 | * |
| 550 | * @param[in] rgb_input_ptr Pointer to the source image. Supported Format: U8 |
| 551 | * @param[in] rgb_input_stride_x Stride of the source image in X dimension (in bytes) |
| 552 | * @param[in] rgb_input_step_x input_stride_x * number of elements along X processed per workitem(in bytes) |
| 553 | * @param[in] rgb_input_stride_y Stride of the source image in Y dimension (in bytes) |
| 554 | * @param[in] rgb_input_step_y rgb_input_stride_y * number of elements along Y processed per workitem(in bytes) |
| 555 | * @param[in] rgb_input_offset_first_element_in_bytes The offset of the first element in the source image |
| 556 | * @param[out] luma_output_ptr Pointer to the destination luma channel. Supported Format: U8 |
| 557 | * @param[in] luma_output_stride_x Stride of the destination luma channel in X dimension (in bytes) |
| 558 | * @param[in] luma_output_step_x luma_output_stride_x * number of elements along X processed per workitem(in bytes) |
| 559 | * @param[in] luma_output_stride_y Stride of the destination image luma channel in Y dimension (in bytes) |
| 560 | * @param[in] luma_output_step_y luma_output_stride_y * number of elements along Y processed per workitem(in bytes) |
| 561 | * @param[in] luma_output_offset_first_element_in_bytes The offset of the first element in the destination luma channel |
| 562 | * @param[out] u_output_ptr Pointer to the destination U channel. Supported Format: U8 |
| 563 | * @param[in] u_output_stride_x Stride of the destination U channel in X dimension (in bytes) |
| 564 | * @param[in] u_output_step_x u_output_stride_x * number of elements along X processed per workitem(in bytes) |
| 565 | * @param[in] u_output_stride_y Stride of the destination image U channel in Y dimension (in bytes) |
| 566 | * @param[in] u_output_step_y u_output_stride_y * number of elements along Y processed per workitem(in bytes) |
| 567 | * @param[in] u_output_offset_first_element_in_bytes The offset of the first element in the destination U channel |
| 568 | * @param[out] v_output_ptr Pointer to the destination V channel. Supported Format: U8 |
| 569 | * @param[in] v_output_stride_x Stride of the destination V channel in X dimension (in bytes) |
| 570 | * @param[in] v_output_step_x v_output_stride_x * number of elements along X processed per workitem(in bytes) |
| 571 | * @param[in] v_output_stride_y Stride of the destination image V channel in Y dimension (in bytes) |
| 572 | * @param[in] v_output_step_y v_output_stride_y * number of elements along Y processed per workitem(in bytes) |
| 573 | * @param[in] v_output_offset_first_element_in_bytes The offset of the first element in the destination V channel |
| 574 | * |
| 575 | */ |
| 576 | __kernel void RGB888_to_YUV444_bt709( |
| 577 | IMAGE_DECLARATION(rgb_input), |
| 578 | IMAGE_DECLARATION(luma_output), |
| 579 | IMAGE_DECLARATION(u_output), |
| 580 | IMAGE_DECLARATION(v_output)) |
| 581 | { |
| 582 | // handle 4 pixels every time |
| 583 | Image in_rgb = CONVERT_TO_IMAGE_STRUCT(rgb_input); |
| 584 | Image out_y = CONVERT_TO_IMAGE_STRUCT(luma_output); |
| 585 | Image out_u = CONVERT_TO_IMAGE_STRUCT(u_output); |
| 586 | Image out_v = CONVERT_TO_IMAGE_STRUCT(v_output); |
| 587 | |
| 588 | // Read 4 pixel |
| 589 | uchar16 rgb_0 = vload16(0, in_rgb.ptr); |
| 590 | uchar4 r_0 = (uchar4)(rgb_0.s0, rgb_0.s3, rgb_0.s6, rgb_0.s9); |
| 591 | uchar4 g_0 = (uchar4)(rgb_0.s1, rgb_0.s4, rgb_0.s7, rgb_0.sa); |
| 592 | uchar4 b_0 = (uchar4)(rgb_0.s2, rgb_0.s5, rgb_0.s8, rgb_0.sb); |
| 593 | |
| 594 | float4 f_y = (float4)(0.0000f) + (float4)(0.2126f) * convert_float4(r_0) + (float4)(0.7152f) * convert_float4(g_0) + (float4)(0.0722f) * convert_float4(b_0); |
| 595 | float4 f_u = (float4)(0.0000f) - (float4)(0.1146f) * convert_float4(r_0) - (float4)(0.3854f) * convert_float4(g_0) + (float4)(0.5000f) * convert_float4(b_0); |
| 596 | float4 f_v = (float4)(0.0000f) + (float4)(0.5000f) * convert_float4(r_0) - (float4)(0.4542f) * convert_float4(g_0) - (float4)(0.0458f) * convert_float4(b_0); |
| 597 | |
| 598 | short4 i_y = convert_short4_rtz(f_y); |
| 599 | short4 i_u = convert_short4_rtz(f_u) + (short4)(128); |
| 600 | short4 i_v = convert_short4_rtz(f_v) + (short4)(128); |
| 601 | |
| 602 | uchar4 luma_0 = convert_uchar4(max((short4)(0), min(i_y, (short4)(255)))); |
| 603 | vstore4(luma_0, 0, out_y.ptr); |
| 604 | |
| 605 | uchar4 cb_0 = convert_uchar4(max((short4)(0), min(i_u, (short4)(255)))); |
| 606 | uchar4 cr_0 = convert_uchar4(max((short4)(0), min(i_v, (short4)(255)))); |
| 607 | vstore4(cb_0, 0, out_u.ptr); |
| 608 | vstore4(cr_0, 0, out_v.ptr); |
| 609 | } |
| 610 | |
| 611 | /** Convert a RGB image to IYUV using BT709 color space |
| 612 | * |
| 613 | * Global Workgroup Size [ DIV_CEIL(width, 2), height ] |
| 614 | * No offset. |
| 615 | * |
| 616 | * @param[in] rgb_input_ptr Pointer to the source image. Supported Format: U8 |
| 617 | * @param[in] rgb_input_stride_x Stride of the source image in X dimension (in bytes) |
| 618 | * @param[in] rgb_input_step_x input_stride_x * number of elements along X processed per workitem(in bytes) |
| 619 | * @param[in] rgb_input_stride_y Stride of the source image in Y dimension (in bytes) |
| 620 | * @param[in] rgb_input_step_y rgb_input_stride_y * number of elements along Y processed per workitem(in bytes) |
| 621 | * @param[in] rgb_input_offset_first_element_in_bytes The offset of the first element in the source image |
| 622 | * @param[out] luma_output_ptr Pointer to the destination luma channel. Supported Format: U8 |
| 623 | * @param[in] luma_output_stride_x Stride of the destination luma channel in X dimension (in bytes) |
| 624 | * @param[in] luma_output_step_x luma_output_stride_x * number of elements along X processed per workitem(in bytes) |
| 625 | * @param[in] luma_output_stride_y Stride of the destination image luma channel in Y dimension (in bytes) |
| 626 | * @param[in] luma_output_step_y luma_output_stride_y * number of elements along Y processed per workitem(in bytes) |
| 627 | * @param[in] luma_output_offset_first_element_in_bytes The offset of the first element in the destination luma channel |
| 628 | * @param[out] u_output_ptr Pointer to the destination U channel. Supported Format: U8 |
| 629 | * @param[in] u_output_stride_x Stride of the destination U channel in X dimension (in bytes) |
| 630 | * @param[in] u_output_step_x u_output_stride_x * number of elements along X processed per workitem(in bytes) |
| 631 | * @param[in] u_output_stride_y Stride of the destination image U channel in Y dimension (in bytes) |
| 632 | * @param[in] u_output_step_y u_output_stride_y * number of elements along Y processed per workitem(in bytes) |
| 633 | * @param[in] u_output_offset_first_element_in_bytes The offset of the first element in the destination U channel |
| 634 | * @param[out] v_output_ptr Pointer to the destination V channel. Supported Format: U8 |
| 635 | * @param[in] v_output_stride_x Stride of the destination V channel in X dimension (in bytes) |
| 636 | * @param[in] v_output_step_x v_output_stride_x * number of elements along X processed per workitem(in bytes) |
| 637 | * @param[in] v_output_stride_y Stride of the destination V channel in Y dimension (in bytes) |
| 638 | * @param[in] v_output_step_y v_output_stride_y * number of elements along Y processed per workitem(in bytes) |
| 639 | * @param[in] v_output_offset_first_element_in_bytes The offset of the first element in the destination V channel |
| 640 | * |
| 641 | */ |
| 642 | __kernel void RGB888_to_IYUV_bt709( |
| 643 | IMAGE_DECLARATION(rgb_input), |
| 644 | IMAGE_DECLARATION(luma_output), |
| 645 | IMAGE_DECLARATION(u_output), |
| 646 | IMAGE_DECLARATION(v_output)) |
| 647 | { |
| 648 | // handle 4 pixels every time, two lines, each line for 2 pixels |
| 649 | Image in_rgb = CONVERT_TO_IMAGE_STRUCT(rgb_input); |
| 650 | Image out_y = CONVERT_TO_IMAGE_STRUCT(luma_output); |
| 651 | Image out_u = CONVERT_TO_IMAGE_STRUCT(u_output); |
| 652 | Image out_v = CONVERT_TO_IMAGE_STRUCT(v_output); |
| 653 | |
| 654 | // Read 2 pixel of the first line |
| 655 | uchar8 rgb_0 = vload8(0, in_rgb.ptr); |
| 656 | uchar2 r_0 = (uchar2)(rgb_0.s0, rgb_0.s3); |
| 657 | uchar2 g_0 = (uchar2)(rgb_0.s1, rgb_0.s4); |
| 658 | uchar2 b_0 = (uchar2)(rgb_0.s2, rgb_0.s5); |
| 659 | |
| 660 | float2 f_y = (float2)(0.0000f) + (float2)(0.2126f) * convert_float2(r_0) + (float2)(0.7152f) * convert_float2(g_0) + (float2)(0.0722f) * convert_float2(b_0); |
| 661 | float2 f_u = (float2)(0.0000f) - (float2)(0.1146f) * convert_float2(r_0) - (float2)(0.3854f) * convert_float2(g_0) + (float2)(0.5000f) * convert_float2(b_0); |
| 662 | float2 f_v = (float2)(0.0000f) + (float2)(0.5000f) * convert_float2(r_0) - (float2)(0.4542f) * convert_float2(g_0) - (float2)(0.0458f) * convert_float2(b_0); |
| 663 | |
| 664 | short2 i_y = convert_short2_rtz(f_y); |
| 665 | short2 i_u = convert_short2_rtz(f_u) + (short2)(128); |
| 666 | short2 i_v = convert_short2_rtz(f_v) + (short2)(128); |
| 667 | |
| 668 | uchar2 luma_0 = convert_uchar2(max((short2)(0), min(i_y, (short2)(255)))); |
| 669 | vstore2(luma_0, 0, out_y.ptr); |
| 670 | |
| 671 | uchar2 cb_0 = convert_uchar2(max((short2)(0), min(i_u, (short2)(255)))); |
| 672 | uchar2 cr_0 = convert_uchar2(max((short2)(0), min(i_v, (short2)(255)))); |
| 673 | |
| 674 | // Read 2 pixel of the second line |
| 675 | uchar8 rgb_1 = vload8(0, in_rgb.ptr + rgb_input_stride_y); |
| 676 | uchar2 r_1 = (uchar2)(rgb_1.s0, rgb_1.s3); |
| 677 | uchar2 g_1 = (uchar2)(rgb_1.s1, rgb_1.s4); |
| 678 | uchar2 b_1 = (uchar2)(rgb_1.s2, rgb_1.s5); |
| 679 | |
| 680 | f_y = (float2)(0.0000f) + (float2)(0.2126f) * convert_float2(r_1) + (float2)(0.7152f) * convert_float2(g_1) + (float2)(0.0722f) * convert_float2(b_1); |
| 681 | f_u = (float2)(0.0000f) - (float2)(0.1146f) * convert_float2(r_1) - (float2)(0.3854f) * convert_float2(g_1) + (float2)(0.5000f) * convert_float2(b_1); |
| 682 | f_v = (float2)(0.0000f) + (float2)(0.5000f) * convert_float2(r_1) - (float2)(0.4542f) * convert_float2(g_1) - (float2)(0.0458f) * convert_float2(b_1); |
| 683 | |
| 684 | i_y = convert_short2_rtz(f_y); |
| 685 | i_u = convert_short2_rtz(f_u) + (short2)(128); |
| 686 | i_v = convert_short2_rtz(f_v) + (short2)(128); |
| 687 | |
| 688 | uchar2 luma_1 = convert_uchar2(max((short2)(0), min(i_y, (short2)(255)))); |
| 689 | vstore2(luma_1, 0, out_y.ptr + luma_output_stride_y); |
| 690 | |
| 691 | uchar2 cb_1 = convert_uchar2(max((short2)(0), min(i_u, (short2)(255)))); |
| 692 | uchar2 cr_1 = convert_uchar2(max((short2)(0), min(i_v, (short2)(255)))); |
| 693 | uchar2 cbcr = (uchar2)(((cb_0.s0 + cb_0.s1 + cb_1.s0 + cb_1.s1) / 4), |
| 694 | ((cr_0.s0 + cr_0.s1 + cr_1.s0 + cr_1.s1) / 4)); |
| 695 | *out_u.ptr = cbcr.x; |
| 696 | *out_v.ptr = cbcr.y; |
| 697 | } |
| 698 | |
| 699 | /** Convert a RGBA image to YUV444 using BT709 color space |
| 700 | * |
| 701 | * Global Workgroup Size [ DIV_CEIL(width, 4), height ] |
| 702 | * No offset. |
| 703 | * |
| 704 | * @param[in] rgba_input_ptr Pointer to the source image. Supported Format: U8 |
| 705 | * @param[in] rgba_input_stride_x Stride of the source image in X dimension (in bytes) |
| 706 | * @param[in] rgba_input_step_x input_stride_x * number of elements along X processed per workitem(in bytes) |
| 707 | * @param[in] rgba_input_stride_y Stride of the source image in Y dimension (in bytes) |
| 708 | * @param[in] rgba_input_step_y rgb_input_stride_y * number of elements along Y processed per workitem(in bytes) |
| 709 | * @param[in] rgba_input_offset_first_element_in_bytes The offset of the first element in the source image |
| 710 | * @param[out] luma_output_ptr Pointer to the destination luma channel. Supported Format: U8 |
| 711 | * @param[in] luma_output_stride_x Stride of the destination luma channel in X dimension (in bytes) |
| 712 | * @param[in] luma_output_step_x luma_output_stride_x * number of elements along X processed per workitem(in bytes) |
| 713 | * @param[in] luma_output_stride_y Stride of the destination image luma channel in Y dimension (in bytes) |
| 714 | * @param[in] luma_output_step_y luma_output_stride_y * number of elements along Y processed per workitem(in bytes) |
| 715 | * @param[in] luma_output_offset_first_element_in_bytes The offset of the first element in the destination luma channel |
| 716 | * @param[out] u_output_ptr Pointer to the destination U channel. Supported Format: U8 |
| 717 | * @param[in] u_output_stride_x Stride of the destination U channel in X dimension (in bytes) |
| 718 | * @param[in] u_output_step_x u_output_stride_x * number of elements along X processed per workitem(in bytes) |
| 719 | * @param[in] u_output_stride_y Stride of the destination image U channel in Y dimension (in bytes) |
| 720 | * @param[in] u_output_step_y u_output_stride_y * number of elements along Y processed per workitem(in bytes) |
| 721 | * @param[in] u_output_offset_first_element_in_bytes The offset of the first element in the destination U channel |
| 722 | * @param[out] v_output_ptr Pointer to the destination V channel. Supported Format: U8 |
| 723 | * @param[in] v_output_stride_x Stride of the destination V channel in X dimension (in bytes) |
| 724 | * @param[in] v_output_step_x v_output_stride_x * number of elements along X processed per workitem(in bytes) |
| 725 | * @param[in] v_output_stride_y Stride of the destination image V channel in Y dimension (in bytes) |
| 726 | * @param[in] v_output_step_y v_output_stride_y * number of elements along Y processed per workitem(in bytes) |
| 727 | * @param[in] v_output_offset_first_element_in_bytes The offset of the first element in the destination V channel |
| 728 | * |
| 729 | */ |
| 730 | __kernel void RGBA8888_to_YUV444_bt709( |
| 731 | IMAGE_DECLARATION(rgba_input), |
| 732 | IMAGE_DECLARATION(luma_output), |
| 733 | IMAGE_DECLARATION(u_output), |
| 734 | IMAGE_DECLARATION(v_output)) |
| 735 | { |
| 736 | // handle 4 pixels every time |
| 737 | Image in_rgba = CONVERT_TO_IMAGE_STRUCT(rgba_input); |
| 738 | Image out_y = CONVERT_TO_IMAGE_STRUCT(luma_output); |
| 739 | Image out_u = CONVERT_TO_IMAGE_STRUCT(u_output); |
| 740 | Image out_v = CONVERT_TO_IMAGE_STRUCT(v_output); |
| 741 | |
| 742 | // Read 4 pixel |
| 743 | uchar16 rgb_0 = vload16(0, in_rgba.ptr); |
| 744 | uchar4 r_0 = (uchar4)(rgb_0.s0, rgb_0.s4, rgb_0.s8, rgb_0.sc); |
| 745 | uchar4 g_0 = (uchar4)(rgb_0.s1, rgb_0.s5, rgb_0.s9, rgb_0.sd); |
| 746 | uchar4 b_0 = (uchar4)(rgb_0.s2, rgb_0.s6, rgb_0.sa, rgb_0.se); |
| 747 | |
| 748 | float4 f_y = (float4)(0.0000f) + (float4)(0.2126f) * convert_float4(r_0) + (float4)(0.7152f) * convert_float4(g_0) + (float4)(0.0722f) * convert_float4(b_0); |
| 749 | float4 f_u = (float4)(0.0000f) - (float4)(0.1146f) * convert_float4(r_0) - (float4)(0.3854f) * convert_float4(g_0) + (float4)(0.5000f) * convert_float4(b_0); |
| 750 | float4 f_v = (float4)(0.0000f) + (float4)(0.5000f) * convert_float4(r_0) - (float4)(0.4542f) * convert_float4(g_0) - (float4)(0.0458f) * convert_float4(b_0); |
| 751 | |
| 752 | short4 i_y = convert_short4(f_y); |
| 753 | short4 i_u = convert_short4(f_u) + (short4)(128); |
| 754 | short4 i_v = convert_short4(f_v) + (short4)(128); |
| 755 | |
| 756 | uchar4 luma_0 = convert_uchar4_sat(max((short4)(0), min(i_y, (short4)(255)))); |
| 757 | vstore4(luma_0, 0, out_y.ptr); |
| 758 | |
| 759 | uchar4 cb_0 = convert_uchar4_sat(max((short4)(0), min(i_u, (short4)(255)))); |
| 760 | uchar4 cr_0 = convert_uchar4_sat(max((short4)(0), min(i_v, (short4)(255)))); |
| 761 | vstore4(cb_0, 0, out_u.ptr); |
| 762 | vstore4(cr_0, 0, out_v.ptr); |
| 763 | } |
| 764 | |
| 765 | /** Convert a RGBA image to NV12 using BT709 color space |
| 766 | * |
| 767 | * Global Workgroup Size [ DIV_CEIL(width, 2), height ] |
| 768 | * No offset. |
| 769 | * |
| 770 | * @param[in] input_ptr Pointer to the source image. Supported Format: U8 |
| 771 | * @param[in] input_stride_x Stride of the source image in X dimension (in bytes) |
| 772 | * @param[in] input_step_x input_stride_x * number of elements along X processed per workitem(in bytes) |
| 773 | * @param[in] input_stride_y Stride of the source image in Y dimension (in bytes) |
| 774 | * @param[in] input_step_y input_stride_y * number of elements along Y processed per workitem(in bytes) |
| 775 | * @param[in] input_offset_first_element_in_bytes The offset of the first element in the source image |
| 776 | * @param[out] luma_output_ptr Pointer to the destination luma channel. Supported Format: U8 |
| 777 | * @param[in] luma_output_stride_x Stride of the destination luma channel in X dimension (in bytes) |
| 778 | * @param[in] luma_output_step_x luma_output_stride_x * number of elements along X processed per workitem(in bytes) |
| 779 | * @param[in] luma_output_stride_y Stride of the destination image luma channel in Y dimension (in bytes) |
| 780 | * @param[in] luma_output_step_y luma_output_stride_y * number of elements along Y processed per workitem(in bytes) |
| 781 | * @param[in] luma_output_offset_first_element_in_bytes The offset of the first element in the destination image luma channel |
| 782 | * @param[out] uv_output_ptr Pointer to the destination uv channel. Supported Format: U8 |
| 783 | * @param[in] uv_output_stride_x Stride of the destination uv channel in X dimension (in bytes) |
| 784 | * @param[in] uv_output_step_x uv_output_stride_x * number of elements along X processed per workitem(in bytes) |
| 785 | * @param[in] uv_output_stride_y Stride of the destination image uv channel in Y dimension (in bytes) |
| 786 | * @param[in] uv_output_step_y uv_output_stride_y * number of elements along Y processed per workitem(in bytes) |
| 787 | * @param[in] uv_output_offset_first_element_in_bytes The offset of the first element in the destination image uv channel |
| 788 | * |
| 789 | */ |
| 790 | __kernel void RGBA8888_to_NV12_bt709( |
| 791 | IMAGE_DECLARATION(input), |
| 792 | IMAGE_DECLARATION(luma_output), |
| 793 | IMAGE_DECLARATION(uv_output)) |
| 794 | { |
| 795 | Image in = CONVERT_TO_IMAGE_STRUCT(input); |
| 796 | Image out_y = CONVERT_TO_IMAGE_STRUCT(luma_output); |
| 797 | Image out_uv = CONVERT_TO_IMAGE_STRUCT(uv_output); |
| 798 | |
| 799 | // Read 2 pixel of the first line |
| 800 | uchar8 rgb_0 = vload8(0, in.ptr); |
| 801 | uchar2 r_0 = (uchar2)(rgb_0.s0, rgb_0.s4); |
| 802 | uchar2 g_0 = (uchar2)(rgb_0.s1, rgb_0.s5); |
| 803 | uchar2 b_0 = (uchar2)(rgb_0.s2, rgb_0.s6); |
| 804 | |
| 805 | float2 f_y = (float2)(0.0000f) + (float2)(0.2126f) * convert_float2(r_0) + (float2)(0.7152f) * convert_float2(g_0) + (float2)(0.0722f) * convert_float2(b_0); |
| 806 | float2 f_u = (float2)(0.0000f) - (float2)(0.1146f) * convert_float2(r_0) - (float2)(0.3854f) * convert_float2(g_0) + (float2)(0.5000f) * convert_float2(b_0); |
| 807 | float2 f_v = (float2)(0.0000f) + (float2)(0.5000f) * convert_float2(r_0) - (float2)(0.4542f) * convert_float2(g_0) - (float2)(0.0458f) * convert_float2(b_0); |
| 808 | |
| 809 | short2 i_y = convert_short2_rtz(f_y); |
| 810 | short2 i_u = convert_short2_rtz(f_u) + (short2)(128); |
| 811 | short2 i_v = convert_short2_rtz(f_v) + (short2)(128); |
| 812 | |
| 813 | uchar2 luma_0 = convert_uchar2(max((short2)(0), min(i_y, (short2)(255)))); |
| 814 | vstore2(luma_0, 0, out_y.ptr); |
| 815 | |
| 816 | uchar2 cb_0 = convert_uchar2(max((short2)(0), min(i_u, (short2)(255)))); |
| 817 | uchar2 cr_0 = convert_uchar2(max((short2)(0), min(i_v, (short2)(255)))); |
| 818 | |
| 819 | // Read 2 pixel of the second line |
| 820 | uchar8 rgb_1 = vload8(0, in.ptr + input_stride_y); |
| 821 | uchar2 r_1 = (uchar2)(rgb_1.s0, rgb_1.s4); |
| 822 | uchar2 g_1 = (uchar2)(rgb_1.s1, rgb_1.s5); |
| 823 | uchar2 b_1 = (uchar2)(rgb_1.s2, rgb_1.s6); |
| 824 | |
| 825 | f_y = (float2)(0.0000f) + (float2)(0.2126f) * convert_float2(r_1) + (float2)(0.7152f) * convert_float2(g_1) + (float2)(0.0722f) * convert_float2(b_1); |
| 826 | f_u = (float2)(0.0000f) - (float2)(0.1146f) * convert_float2(r_1) - (float2)(0.3854f) * convert_float2(g_1) + (float2)(0.5000f) * convert_float2(b_1); |
| 827 | f_v = (float2)(0.0000f) + (float2)(0.5000f) * convert_float2(r_1) - (float2)(0.4542f) * convert_float2(g_1) - (float2)(0.0458f) * convert_float2(b_1); |
| 828 | |
| 829 | i_y = convert_short2_rtz(f_y); |
| 830 | i_u = convert_short2_rtz(f_u) + (short2)(128); |
| 831 | i_v = convert_short2_rtz(f_v) + (short2)(128); |
| 832 | |
| 833 | uchar2 luma_1 = convert_uchar2(max((short2)(0), min(i_y, (short2)(255)))); |
| 834 | vstore2(luma_1, 0, out_y.ptr + luma_output_stride_y); |
| 835 | |
| 836 | uchar2 cb_1 = convert_uchar2(max((short2)(0), min(i_u, (short2)(255)))); |
| 837 | uchar2 cr_1 = convert_uchar2(max((short2)(0), min(i_v, (short2)(255)))); |
| 838 | uchar2 cbcr = (uchar2)(((cb_0.s0 + cb_0.s1 + cb_1.s0 + cb_1.s1) / 4), |
| 839 | ((cr_0.s0 + cr_0.s1 + cr_1.s0 + cr_1.s1) / 4)); |
| 840 | vstore2(cbcr, 0, out_uv.ptr); |
| 841 | } |
| 842 | |
| 843 | /** Convert a RGBA image to IYUV using BT709 color space |
| 844 | * |
| 845 | * Global Workgroup Size [ DIV_CEIL(width, 2), height ] |
| 846 | * No offset. |
| 847 | * |
| 848 | * @param[in] rgba_input_ptr Pointer to the source image. Supported Format: U8 |
| 849 | * @param[in] rgba_input_stride_x Stride of the source image in X dimension (in bytes) |
| 850 | * @param[in] rgba_input_step_x rgba_input_stride_x * number of elements along X processed per workitem(in bytes) |
| 851 | * @param[in] rgba_input_stride_y Stride of the source image in Y dimension (in bytes) |
| 852 | * @param[in] rgba_input_step_y rgba_input_stride_y * number of elements along Y processed per workitem(in bytes) |
| 853 | * @param[in] rgba_input_offset_first_element_in_bytes The offset of the first element in the source image |
| 854 | * @param[out] luma_output_ptr Pointer to the destination luma channel. Supported Format: U8 |
| 855 | * @param[in] luma_output_stride_x Stride of the destination luma channel in X dimension (in bytes) |
| 856 | * @param[in] luma_output_step_x luma_output_stride_x * number of elements along X processed per workitem(in bytes) |
| 857 | * @param[in] luma_output_stride_y Stride of the destination image luma channel in Y dimension (in bytes) |
| 858 | * @param[in] luma_output_step_y luma_output_stride_y * number of elements along Y processed per workitem(in bytes) |
| 859 | * @param[in] luma_output_offset_first_element_in_bytes The offset of the first element in the destination luma channel |
| 860 | * @param[out] u_output_ptr Pointer to the destination U channel. Supported Format: U8 |
| 861 | * @param[in] u_output_stride_x Stride of the destination U channel in X dimension (in bytes) |
| 862 | * @param[in] u_output_step_x u_output_stride_x * number of elements along X processed per workitem(in bytes) |
| 863 | * @param[in] u_output_stride_y Stride of the destination image U channel in Y dimension (in bytes) |
| 864 | * @param[in] u_output_step_y u_output_stride_y * number of elements along Y processed per workitem(in bytes) |
| 865 | * @param[in] u_output_offset_first_element_in_bytes The offset of the first element in the destination U channel |
| 866 | * @param[out] v_output_ptr Pointer to the destination V channel. Supported Format: U8 |
| 867 | * @param[in] v_output_stride_x Stride of the destination V channel in X dimension (in bytes) |
| 868 | * @param[in] v_output_step_x v_output_stride_x * number of elements along X processed per workitem(in bytes) |
| 869 | * @param[in] v_output_stride_y Stride of the destination V channel in Y dimension (in bytes) |
| 870 | * @param[in] v_output_step_y v_output_stride_y * number of elements along Y processed per workitem(in bytes) |
| 871 | * @param[in] v_output_offset_first_element_in_bytes The offset of the first element in the destination V channel |
| 872 | * |
| 873 | */ |
| 874 | __kernel void RGBA8888_to_IYUV_bt709( |
| 875 | IMAGE_DECLARATION(rgba_input), |
| 876 | IMAGE_DECLARATION(luma_output), |
| 877 | IMAGE_DECLARATION(u_output), |
| 878 | IMAGE_DECLARATION(v_output)) |
| 879 | { |
| 880 | // handle 4 pixels every time, two lines, each line for 2 pixels |
| 881 | Image in_rgb = CONVERT_TO_IMAGE_STRUCT(rgba_input); |
| 882 | Image out_y = CONVERT_TO_IMAGE_STRUCT(luma_output); |
| 883 | Image out_u = CONVERT_TO_IMAGE_STRUCT(u_output); |
| 884 | Image out_v = CONVERT_TO_IMAGE_STRUCT(v_output); |
| 885 | |
| 886 | // Read 2 pixel of the first line |
| 887 | uchar8 rgb_0 = vload8(0, in_rgb.ptr); |
| 888 | uchar2 r_0 = (uchar2)(rgb_0.s0, rgb_0.s4); |
| 889 | uchar2 g_0 = (uchar2)(rgb_0.s1, rgb_0.s5); |
| 890 | uchar2 b_0 = (uchar2)(rgb_0.s2, rgb_0.s6); |
| 891 | |
| 892 | float2 f_y = (float2)(0.0000f) + (float2)(0.2126f) * convert_float2(r_0) + (float2)(0.7152f) * convert_float2(g_0) + (float2)(0.0722f) * convert_float2(b_0); |
| 893 | float2 f_u = (float2)(0.0000f) - (float2)(0.1146f) * convert_float2(r_0) - (float2)(0.3854f) * convert_float2(g_0) + (float2)(0.5000f) * convert_float2(b_0); |
| 894 | float2 f_v = (float2)(0.0000f) + (float2)(0.5000f) * convert_float2(r_0) - (float2)(0.4542f) * convert_float2(g_0) - (float2)(0.0458f) * convert_float2(b_0); |
| 895 | |
| 896 | short2 i_y = convert_short2_rtz(f_y); |
| 897 | short2 i_u = convert_short2_rtz(f_u) + (short2)(128); |
| 898 | short2 i_v = convert_short2_rtz(f_v) + (short2)(128); |
| 899 | |
| 900 | uchar2 luma_0 = convert_uchar2(max((short2)(0), min(i_y, (short2)(255)))); |
| 901 | vstore2(luma_0, 0, out_y.ptr); |
| 902 | |
| 903 | uchar2 cb_0 = convert_uchar2(max((short2)(0), min(i_u, (short2)(255)))); |
| 904 | uchar2 cr_0 = convert_uchar2(max((short2)(0), min(i_v, (short2)(255)))); |
| 905 | |
| 906 | // Read 2 pixel of the second line |
| 907 | uchar8 rgb_1 = vload8(0, in_rgb.ptr + rgba_input_stride_y); |
| 908 | uchar2 r_1 = (uchar2)(rgb_1.s0, rgb_1.s4); |
| 909 | uchar2 g_1 = (uchar2)(rgb_1.s1, rgb_1.s5); |
| 910 | uchar2 b_1 = (uchar2)(rgb_1.s2, rgb_1.s6); |
| 911 | |
| 912 | f_y = (float2)(0.0000f) + (float2)(0.2126f) * convert_float2(r_1) + (float2)(0.7152f) * convert_float2(g_1) + (float2)(0.0722f) * convert_float2(b_1); |
| 913 | f_u = (float2)(0.0000f) - (float2)(0.1146f) * convert_float2(r_1) - (float2)(0.3854f) * convert_float2(g_1) + (float2)(0.5000f) * convert_float2(b_1); |
| 914 | f_v = (float2)(0.0000f) + (float2)(0.5000f) * convert_float2(r_1) - (float2)(0.4542f) * convert_float2(g_1) - (float2)(0.0458f) * convert_float2(b_1); |
| 915 | |
| 916 | i_y = convert_short2_rtz(f_y); |
| 917 | i_u = convert_short2_rtz(f_u) + (short2)(128); |
| 918 | i_v = convert_short2_rtz(f_v) + (short2)(128); |
| 919 | |
| 920 | uchar2 luma_1 = convert_uchar2(max((short2)(0), min(i_y, (short2)(255)))); |
| 921 | vstore2(luma_1, 0, out_y.ptr + luma_output_stride_y); |
| 922 | |
| 923 | uchar2 cb_1 = convert_uchar2(max((short2)(0), min(i_u, (short2)(255)))); |
| 924 | uchar2 cr_1 = convert_uchar2(max((short2)(0), min(i_v, (short2)(255)))); |
| 925 | uchar2 cbcr = (uchar2)(((cb_0.s0 + cb_0.s1 + cb_1.s0 + cb_1.s1) / 4), |
| 926 | ((cr_0.s0 + cr_0.s1 + cr_1.s0 + cr_1.s1) / 4)); |
| 927 | *out_u.ptr = cbcr.x; |
| 928 | *out_v.ptr = cbcr.y; |
| 929 | } |
| 930 | |
| 931 | /** Convert an NV12 image to RGB8888 |
| 932 | * |
| 933 | * Global Workgroup Size [ DIV_CEIL(width, 4), height ] |
| 934 | * No offset. |
| 935 | * |
| 936 | * @param[in] luma_input_ptr Pointer to the source luma channel. Supported Format: U8 |
| 937 | * @param[in] luma_input_stride_x Stride of the luma image in X dimension (in bytes) |
| 938 | * @param[in] luma_input_step_x luma_input_stride_x * number of elements along X processed per workitem(in bytes) |
| 939 | * @param[in] luma_input_stride_y Stride of the source luma channel in Y dimension (in bytes) |
| 940 | * @param[in] luma_input_step_y luma_input_stride_y * number of elements along Y processed per workitem(in bytes) |
| 941 | * @param[in] luma_input_offset_first_element_in_bytes The offset of the first element in the source image |
| 942 | * @param[in] uv_input_ptr Pointer to the source uv channel. Supported Format: U8 |
| 943 | * @param[in] uv_input_stride_x Stride of the source image uv channel in X dimension (in bytes) |
| 944 | * @param[in] uv_input_step_x uv_input_stride_x * number of elements along X processed per workitem(in bytes) |
| 945 | * @param[in] uv_input_stride_y Stride of the source image in Y dimension (in bytes) |
| 946 | * @param[in] uv_input_step_y uv_input_stride_y * number of elements along Y processed per workitem(in bytes) |
| 947 | * @param[in] uv_input_offset_first_element_in_bytes The offset of the first element in the source image |
| 948 | * @param[out] rgb_output_ptr Pointer to the destination image. Supported Format: U8 |
| 949 | * @param[in] rgb_output_stride_x Stride of the destination image in X dimension (in bytes) |
| 950 | * @param[in] rgb_output_step_x rgb_output_stride_x * number of elements along X processed per workitem(in bytes) |
| 951 | * @param[in] rgb_output_stride_y Stride of the destination image in Y dimension (in bytes) |
| 952 | * @param[in] rgb_output_step_y rgb_output_stride_y * number of elements along Y processed per workitem(in bytes) |
| 953 | * @param[in] rgb_output_offset_first_element_in_bytes The offset of the first element in the destination image |
| 954 | */ |
| 955 | __kernel void NV12_to_RGBA8888_bt709( |
| 956 | IMAGE_DECLARATION(luma_input), |
| 957 | IMAGE_DECLARATION(uv_input), |
| 958 | IMAGE_DECLARATION(rgb_output)) |
| 959 | { |
| 960 | Image in_luma = CONVERT_TO_IMAGE_STRUCT(luma_input); |
| 961 | Image in_uv = CONVERT_TO_IMAGE_STRUCT(uv_input); |
| 962 | Image out_rgb = CONVERT_TO_IMAGE_STRUCT(rgb_output); |
| 963 | |
| 964 | uchar4 luma_0 = vload4(0, in_luma.ptr); |
| 965 | uchar4 luma_1 = vload4(0, in_luma.ptr + luma_input_stride_y); |
| 966 | uchar4 cbcr = vload4(0, in_uv.ptr); |
| 967 | char4 cb = (char4)(cbcr.s0, cbcr.s0, cbcr.s2, cbcr.s2) - (char4)(128); |
| 968 | char4 cr = (char4)(cbcr.s1, cbcr.s1, cbcr.s3, cbcr.s3) - (char4)(128); |
| 969 | |
| 970 | float4 temp0 = (float4)(0.0000f) + (float4)(0.0000f) * convert_float4(cb) + (float4)(1.5748f) * convert_float4(cr); |
| 971 | float4 temp1 = (float4)(0.0000f) - (float4)(0.1873f) * convert_float4(cb) - (float4)(0.4681f) * convert_float4(cr); |
| 972 | float4 temp2 = (float4)(0.0000f) + (float4)(1.8556f) * convert_float4(cb) + (float4)(0.0000f) * convert_float4(cr); |
| 973 | |
| 974 | float4 f_r = convert_float4(luma_0) + temp0; |
| 975 | float4 f_g = convert_float4(luma_0) + temp1; |
| 976 | float4 f_b = convert_float4(luma_0) + temp2; |
| 977 | |
Pablo Tello | 96fc1d6 | 2018-07-17 17:10:59 +0100 | [diff] [blame] | 978 | uchar4 r_0 = convert_uchar4_sat_rtz(f_r); |
| 979 | uchar4 g_0 = convert_uchar4_sat_rtz(f_g); |
| 980 | uchar4 b_0 = convert_uchar4_sat_rtz(f_b); |
Anthony Barbier | 6ff3b19 | 2017-09-04 18:44:23 +0100 | [diff] [blame] | 981 | |
| 982 | uchar8 rgb_0 = (uchar8)(r_0.s0, g_0.s0, b_0.s0, 255, r_0.s1, g_0.s1, b_0.s1, 255); |
| 983 | uchar8 rgb_1 = (uchar8)(r_0.s2, g_0.s2, b_0.s2, 255, r_0.s3, g_0.s3, b_0.s3, 255); |
| 984 | vstore8(rgb_0, 0, out_rgb.ptr); |
| 985 | vstore8(rgb_1, 0, out_rgb.ptr + 8); |
| 986 | |
| 987 | f_r = convert_float4(luma_1) + temp0; |
| 988 | f_g = convert_float4(luma_1) + temp1; |
| 989 | f_b = convert_float4(luma_1) + temp2; |
| 990 | |
Pablo Tello | 96fc1d6 | 2018-07-17 17:10:59 +0100 | [diff] [blame] | 991 | r_0 = convert_uchar4_sat_rtz(f_r); |
| 992 | g_0 = convert_uchar4_sat_rtz(f_g); |
| 993 | b_0 = convert_uchar4_sat_rtz(f_b); |
Anthony Barbier | 6ff3b19 | 2017-09-04 18:44:23 +0100 | [diff] [blame] | 994 | |
| 995 | rgb_0 = (uchar8)(r_0.s0, g_0.s0, b_0.s0, 255, r_0.s1, g_0.s1, b_0.s1, 255); |
| 996 | rgb_1 = (uchar8)(r_0.s2, g_0.s2, b_0.s2, 255, r_0.s3, g_0.s3, b_0.s3, 255); |
| 997 | vstore8(rgb_0, 0, out_rgb.ptr + rgb_output_stride_y); |
| 998 | vstore8(rgb_1, 0, out_rgb.ptr + rgb_output_stride_y + 8); |
| 999 | } |
| 1000 | |
| 1001 | /** Convert an NV12 image to IYUV |
| 1002 | * |
| 1003 | * Global Workgroup Size [ DIV_CEIL(width, 16), height ] |
| 1004 | * No offset. |
| 1005 | * |
| 1006 | * @param[in] luma_input_ptr Pointer to the source luma channel. Supported Format: U8 |
| 1007 | * @param[in] luma_input_stride_x Stride of the luma image in X dimension (in bytes) |
| 1008 | * @param[in] luma_input_step_x luma_input_stride_x * number of elements along X processed per workitem(in bytes) |
| 1009 | * @param[in] luma_input_stride_y Stride of the source luma channel in Y dimension (in bytes) |
| 1010 | * @param[in] luma_input_step_y luma_input_stride_y * number of elements along Y processed per workitem(in bytes) |
| 1011 | * @param[in] luma_input_offset_first_element_in_bytes The offset of the first element in the source image |
| 1012 | * @param[in] uv_input_ptr Pointer to the source uv channel. Supported Format: U8 |
| 1013 | * @param[in] uv_input_stride_x Stride of the source image uv channel in X dimension (in bytes) |
| 1014 | * @param[in] uv_input_step_x uv_input_stride_x * number of elements along X processed per workitem(in bytes) |
| 1015 | * @param[in] uv_input_stride_y Stride of the source image in Y dimension (in bytes) |
| 1016 | * @param[in] uv_input_step_y uv_input_stride_y * number of elements along Y processed per workitem(in bytes) |
| 1017 | * @param[in] uv_input_offset_first_element_in_bytes The offset of the first element in the source image |
| 1018 | * @param[out] luma_output_ptr Pointer to the destination luma channel. Supported Format: U8 |
| 1019 | * @param[in] luma_output_stride_x Stride of the destination luma channel in X dimension (in bytes) |
| 1020 | * @param[in] luma_output_step_x luma_output_stride_x * number of elements along X processed per workitem(in bytes) |
| 1021 | * @param[in] luma_output_stride_y Stride of the destination image luma channel in Y dimension (in bytes) |
| 1022 | * @param[in] luma_output_step_y luma_output_stride_y * number of elements along Y processed per workitem(in bytes) |
| 1023 | * @param[in] luma_output_offset_first_element_in_bytes The offset of the first element in the destination luma channel |
| 1024 | * @param[out] u_output_ptr Pointer to the destination U channel. Supported Format: U8 |
| 1025 | * @param[in] u_output_stride_x Stride of the destination U channel in X dimension (in bytes) |
| 1026 | * @param[in] u_output_step_x u_output_stride_x * number of elements along X processed per workitem(in bytes) |
| 1027 | * @param[in] u_output_stride_y Stride of the destination image U channel in Y dimension (in bytes) |
| 1028 | * @param[in] u_output_step_y u_output_stride_y * number of elements along Y processed per workitem(in bytes) |
| 1029 | * @param[in] u_output_offset_first_element_in_bytes The offset of the first element in the destination U channel |
| 1030 | * @param[out] v_output_ptr Pointer to the destination V channel. Supported Format: U8 |
| 1031 | * @param[in] v_output_stride_x Stride of the destination V channel in X dimension (in bytes) |
| 1032 | * @param[in] v_output_step_x v_output_stride_x * number of elements along X processed per workitem(in bytes) |
| 1033 | * @param[in] v_output_stride_y Stride of the destination V channel in Y dimension (in bytes) |
| 1034 | * @param[in] v_output_step_y v_output_stride_y * number of elements along Y processed per workitem(in bytes) |
| 1035 | * @param[in] v_output_offset_first_element_in_bytes The offset of the first element in the destination V channel |
| 1036 | */ |
| 1037 | __kernel void NV12_to_IYUV_bt709( |
| 1038 | IMAGE_DECLARATION(luma_input), |
| 1039 | IMAGE_DECLARATION(uv_input), |
| 1040 | IMAGE_DECLARATION(luma_output), |
| 1041 | IMAGE_DECLARATION(u_output), |
| 1042 | IMAGE_DECLARATION(v_output)) |
| 1043 | { |
| 1044 | Image in_y = CONVERT_TO_IMAGE_STRUCT(luma_input); |
| 1045 | Image in_uv = CONVERT_TO_IMAGE_STRUCT(uv_input); |
| 1046 | Image out_y = CONVERT_TO_IMAGE_STRUCT(luma_output); |
| 1047 | Image out_u = CONVERT_TO_IMAGE_STRUCT(u_output); |
| 1048 | Image out_v = CONVERT_TO_IMAGE_STRUCT(v_output); |
| 1049 | |
| 1050 | // handle 32 pixels every time, two lines, each line for 16 pixels |
| 1051 | uchar16 luma_0 = vload16(0, in_y.ptr); |
| 1052 | uchar16 luma_1 = vload16(0, in_y.ptr + luma_input_stride_y); |
| 1053 | uchar16 cbcr = vload16(0, in_uv.ptr); |
| 1054 | uchar8 cb = (uchar8)(cbcr.s0, cbcr.s2, cbcr.s4, cbcr.s6, cbcr.s8, cbcr.sa, cbcr.sc, cbcr.se); |
| 1055 | uchar8 cr = (uchar8)(cbcr.s1, cbcr.s3, cbcr.s5, cbcr.s7, cbcr.s9, cbcr.sb, cbcr.sd, cbcr.sf); |
| 1056 | |
| 1057 | vstore16(luma_0, 0, out_y.ptr); |
| 1058 | vstore16(luma_1, 0, out_y.ptr + luma_output_stride_y); |
| 1059 | vstore8(cb, 0, out_u.ptr); |
| 1060 | vstore8(cr, 0, out_v.ptr); |
| 1061 | } |
| 1062 | |
| 1063 | /** Convert an NV12 image to YUV444 |
| 1064 | * |
| 1065 | * Global Workgroup Size [ DIV_CEIL(width, 16), height ] |
| 1066 | * No offset. |
| 1067 | * |
| 1068 | * @param[in] luma_input_ptr Pointer to the source luma channel. Supported Format: U8 |
| 1069 | * @param[in] luma_input_stride_x Stride of the luma image in X dimension (in bytes) |
| 1070 | * @param[in] luma_input_step_x luma_input_stride_x * number of elements along X processed per workitem(in bytes) |
| 1071 | * @param[in] luma_input_stride_y Stride of the source luma channel in Y dimension (in bytes) |
| 1072 | * @param[in] luma_input_step_y luma_input_stride_y * number of elements along Y processed per workitem(in bytes) |
| 1073 | * @param[in] luma_input_offset_first_element_in_bytes The offset of the first element in the source image |
| 1074 | * @param[in] uv_input_ptr Pointer to the source uv channel. Supported Format: U8 |
| 1075 | * @param[in] uv_input_stride_x Stride of the source image uv channel in X dimension (in bytes) |
| 1076 | * @param[in] uv_input_step_x uv_input_stride_x * number of elements along X processed per workitem(in bytes) |
| 1077 | * @param[in] uv_input_stride_y Stride of the source image in Y dimension (in bytes) |
| 1078 | * @param[in] uv_input_step_y uv_input_stride_y * number of elements along Y processed per workitem(in bytes) |
| 1079 | * @param[in] uv_input_offset_first_element_in_bytes The offset of the first element in the source image |
| 1080 | * @param[out] luma_output_ptr Pointer to the destination luma channel. Supported Format: U8 |
| 1081 | * @param[in] luma_output_stride_x Stride of the destination luma channel in X dimension (in bytes) |
| 1082 | * @param[in] luma_output_step_x luma_output_stride_x * number of elements along X processed per workitem(in bytes) |
| 1083 | * @param[in] luma_output_stride_y Stride of the destination image luma channel in Y dimension (in bytes) |
| 1084 | * @param[in] luma_output_step_y luma_output_stride_y * number of elements along Y processed per workitem(in bytes) |
| 1085 | * @param[in] luma_output_offset_first_element_in_bytes The offset of the first element in the destination luma channel |
| 1086 | * @param[out] u_output_ptr Pointer to the destination U channel. Supported Format: U8 |
| 1087 | * @param[in] u_output_stride_x Stride of the destination U channel in X dimension (in bytes) |
| 1088 | * @param[in] u_output_step_x u_output_stride_x * number of elements along X processed per workitem(in bytes) |
| 1089 | * @param[in] u_output_stride_y Stride of the destination image U channel in Y dimension (in bytes) |
| 1090 | * @param[in] u_output_step_y u_output_stride_y * number of elements along Y processed per workitem(in bytes) |
| 1091 | * @param[in] u_output_offset_first_element_in_bytes The offset of the first element in the destination U channel |
| 1092 | * @param[out] v_output_ptr Pointer to the destination V channel. Supported Format: U8 |
| 1093 | * @param[in] v_output_stride_x Stride of the destination V channel in X dimension (in bytes) |
| 1094 | * @param[in] v_output_step_x v_output_stride_x * number of elements along X processed per workitem(in bytes) |
| 1095 | * @param[in] v_output_stride_y Stride of the destination V channel in Y dimension (in bytes) |
| 1096 | * @param[in] v_output_step_y v_output_stride_y * number of elements along Y processed per workitem(in bytes) |
| 1097 | * @param[in] v_output_offset_first_element_in_bytes The offset of the first element in the destination V channel |
| 1098 | */ |
| 1099 | __kernel void NV12_to_YUV444_bt709( |
| 1100 | IMAGE_DECLARATION(luma_input), |
| 1101 | IMAGE_DECLARATION(uv_input), |
| 1102 | IMAGE_DECLARATION(luma_output), |
| 1103 | IMAGE_DECLARATION(u_output), |
| 1104 | IMAGE_DECLARATION(v_output)) |
| 1105 | { |
| 1106 | Image in_y = CONVERT_TO_IMAGE_STRUCT(luma_input); |
| 1107 | Image in_uv = CONVERT_TO_IMAGE_STRUCT(uv_input); |
| 1108 | Image out_y = CONVERT_TO_IMAGE_STRUCT(luma_output); |
| 1109 | Image out_u = CONVERT_TO_IMAGE_STRUCT(u_output); |
| 1110 | Image out_v = CONVERT_TO_IMAGE_STRUCT(v_output); |
| 1111 | |
| 1112 | // handle 32 pixels every time, two lines, each line for 16 pixels |
| 1113 | uchar16 luma_0 = vload16(0, in_y.ptr); |
| 1114 | uchar16 luma_1 = vload16(0, in_y.ptr + luma_input_stride_y); |
| 1115 | uchar16 cbcr = vload16(0, in_uv.ptr); |
| 1116 | uchar16 cb = (uchar16)(cbcr.s0, cbcr.s0, cbcr.s2, cbcr.s2, cbcr.s4, cbcr.s4, cbcr.s6, cbcr.s6, cbcr.s8, cbcr.s8, |
| 1117 | cbcr.sa, cbcr.sa, cbcr.sc, cbcr.sc, cbcr.se, cbcr.se); |
| 1118 | uchar16 cr = (uchar16)(cbcr.s1, cbcr.s1, cbcr.s3, cbcr.s3, cbcr.s5, cbcr.s5, cbcr.s7, cbcr.s7, cbcr.s9, cbcr.s9, |
| 1119 | cbcr.sb, cbcr.sb, cbcr.sd, cbcr.sd, cbcr.sf, cbcr.sf); |
| 1120 | |
| 1121 | vstore16(luma_0, 0, out_y.ptr); |
| 1122 | vstore16(luma_1, 0, out_y.ptr + luma_output_stride_y); |
| 1123 | vstore16(cb, 0, out_u.ptr); |
| 1124 | vstore16(cb, 0, out_u.ptr + u_output_stride_y); |
| 1125 | vstore16(cr, 0, out_v.ptr); |
| 1126 | vstore16(cr, 0, out_v.ptr + v_output_stride_y); |
| 1127 | } |
| 1128 | |
| 1129 | /** Convert an NV21 image to RGB888 |
| 1130 | * |
| 1131 | * Global Workgroup Size [ DIV_CEIL(width, 4), height ] |
| 1132 | * No offset. |
| 1133 | * |
| 1134 | * @param[in] luma_input_ptr Pointer to the source luma channel. Supported Format: U8 |
| 1135 | * @param[in] luma_input_stride_x Stride of the luma image in X dimension (in bytes) |
| 1136 | * @param[in] luma_input_step_x luma_input_stride_x * number of elements along X processed per workitem(in bytes) |
| 1137 | * @param[in] luma_input_stride_y Stride of the source luma channel in Y dimension (in bytes) |
| 1138 | * @param[in] luma_input_step_y luma_input_stride_y * number of elements along Y processed per workitem(in bytes) |
| 1139 | * @param[in] luma_input_offset_first_element_in_bytes The offset of the first element in the source image |
| 1140 | * @param[in] uv_input_ptr Pointer to the source uv channel. Supported Format: U8 |
| 1141 | * @param[in] uv_input_stride_x Stride of the source image uv channel in X dimension (in bytes) |
| 1142 | * @param[in] uv_input_step_x uv_input_stride_x * number of elements along X processed per workitem(in bytes) |
| 1143 | * @param[in] uv_input_stride_y Stride of the source image in Y dimension (in bytes) |
| 1144 | * @param[in] uv_input_step_y uv_input_stride_y * number of elements along Y processed per workitem(in bytes) |
| 1145 | * @param[in] uv_input_offset_first_element_in_bytes The offset of the first element in the source image |
| 1146 | * @param[out] rgb_output_ptr Pointer to the destination image. Supported Format: U8 |
| 1147 | * @param[in] rgb_output_stride_x Stride of the destination image in X dimension (in bytes) |
| 1148 | * @param[in] rgb_output_step_x rgb_output_stride_x * number of elements along X processed per workitem(in bytes) |
| 1149 | * @param[in] rgb_output_stride_y Stride of the destination image in Y dimension (in bytes) |
| 1150 | * @param[in] rgb_output_step_y output_stride_y * number of elements along Y processed per workitem(in bytes) |
| 1151 | * @param[in] rgb_output_offset_first_element_in_bytes The offset of the first element in the destination image |
| 1152 | */ |
| 1153 | __kernel void NV21_to_RGB888_bt709( |
| 1154 | IMAGE_DECLARATION(luma_input), |
| 1155 | IMAGE_DECLARATION(uv_input), |
| 1156 | IMAGE_DECLARATION(rgb_output)) |
| 1157 | { |
| 1158 | Image in_y = CONVERT_TO_IMAGE_STRUCT(luma_input); |
| 1159 | Image in_uv = CONVERT_TO_IMAGE_STRUCT(uv_input); |
| 1160 | Image out_rgb = CONVERT_TO_IMAGE_STRUCT(rgb_output); |
| 1161 | |
| 1162 | // handle 8 pixels every time, two lines, each line for 4 pixels |
| 1163 | uchar4 luma_0 = vload4(0, in_y.ptr); |
| 1164 | uchar4 luma_1 = vload4(0, in_y.ptr + luma_input_stride_y); |
| 1165 | uchar4 cbcr = vload4(0, in_uv.ptr); |
| 1166 | char4 cr = (char4)(cbcr.s0, cbcr.s0, cbcr.s2, cbcr.s2) - (char4)(128); |
| 1167 | char4 cb = (char4)(cbcr.s1, cbcr.s1, cbcr.s3, cbcr.s3) - (char4)(128); |
| 1168 | |
| 1169 | float4 temp0 = (float4)(0.0000f) + (float4)(0.0000f) * convert_float4(cb) + (float4)(1.5748f) * convert_float4(cr); |
| 1170 | float4 temp1 = (float4)(0.0000f) - (float4)(0.1873f) * convert_float4(cb) - (float4)(0.4681f) * convert_float4(cr); |
| 1171 | float4 temp2 = (float4)(0.0000f) + (float4)(1.8556f) * convert_float4(cb) + (float4)(0.0000f) * convert_float4(cr); |
| 1172 | |
| 1173 | float4 f_r = convert_float4(luma_0) + temp0; |
| 1174 | float4 f_g = convert_float4(luma_0) + temp1; |
| 1175 | float4 f_b = convert_float4(luma_0) + temp2; |
| 1176 | |
Pablo Tello | 96fc1d6 | 2018-07-17 17:10:59 +0100 | [diff] [blame] | 1177 | uchar4 r_0 = convert_uchar4_sat_rtz(f_r); |
| 1178 | uchar4 g_0 = convert_uchar4_sat_rtz(f_g); |
| 1179 | uchar4 b_0 = convert_uchar4_sat_rtz(f_b); |
Anthony Barbier | 6ff3b19 | 2017-09-04 18:44:23 +0100 | [diff] [blame] | 1180 | |
| 1181 | uchar8 rgb_0 = (uchar8)(r_0.s0, g_0.s0, b_0.s0, r_0.s1, g_0.s1, b_0.s1, r_0.s2, g_0.s2); |
| 1182 | uchar4 rgb_1 = (uchar4)(b_0.s2, r_0.s3, g_0.s3, b_0.s3); |
| 1183 | vstore8(rgb_0, 0, out_rgb.ptr); |
| 1184 | vstore4(rgb_1, 0, out_rgb.ptr + 8); |
| 1185 | |
| 1186 | f_r = convert_float4(luma_1) + temp0; |
| 1187 | f_g = convert_float4(luma_1) + temp1; |
| 1188 | f_b = convert_float4(luma_1) + temp2; |
| 1189 | |
Pablo Tello | 96fc1d6 | 2018-07-17 17:10:59 +0100 | [diff] [blame] | 1190 | r_0 = convert_uchar4_sat_rtz(f_r); |
| 1191 | g_0 = convert_uchar4_sat_rtz(f_g); |
| 1192 | b_0 = convert_uchar4_sat_rtz(f_b); |
Anthony Barbier | 6ff3b19 | 2017-09-04 18:44:23 +0100 | [diff] [blame] | 1193 | |
| 1194 | rgb_0 = (uchar8)(r_0.s0, g_0.s0, b_0.s0, r_0.s1, g_0.s1, b_0.s1, r_0.s2, g_0.s2); |
| 1195 | rgb_1 = (uchar4)(b_0.s2, r_0.s3, g_0.s3, b_0.s3); |
| 1196 | vstore8(rgb_0, 0, out_rgb.ptr + rgb_output_stride_y); |
| 1197 | vstore4(rgb_1, 0, out_rgb.ptr + rgb_output_stride_y + 8); |
| 1198 | } |
| 1199 | |
| 1200 | /** Convert an NV12 image to RGB8888 |
| 1201 | * |
| 1202 | * Global Workgroup Size [ DIV_CEIL(width, 4), height ] |
| 1203 | * No offset. |
| 1204 | * |
| 1205 | * @param[in] luma_input_ptr Pointer to the source luma channel. Supported Format: U8 |
| 1206 | * @param[in] luma_input_stride_x Stride of the luma image in X dimension (in bytes) |
| 1207 | * @param[in] luma_input_step_x luma_input_stride_x * number of elements along X processed per workitem(in bytes) |
| 1208 | * @param[in] luma_input_stride_y Stride of the source luma channel in Y dimension (in bytes) |
| 1209 | * @param[in] luma_input_step_y luma_input_stride_y * number of elements along Y processed per workitem(in bytes) |
| 1210 | * @param[in] luma_input_offset_first_element_in_bytes The offset of the first element in the source image |
| 1211 | * @param[in] uv_input_ptr Pointer to the source uv channel. Supported Format: U8 |
| 1212 | * @param[in] uv_input_stride_x Stride of the source image uv channel in X dimension (in bytes) |
| 1213 | * @param[in] uv_input_step_x uv_input_stride_x * number of elements along X processed per workitem(in bytes) |
| 1214 | * @param[in] uv_input_stride_y Stride of the source image in Y dimension (in bytes) |
| 1215 | * @param[in] uv_input_step_y uv_input_stride_y * number of elements along Y processed per workitem(in bytes) |
| 1216 | * @param[in] uv_input_offset_first_element_in_bytes The offset of the first element in the source image |
| 1217 | * @param[out] rgba_output_ptr Pointer to the destination image. Supported Format: U8 |
| 1218 | * @param[in] rgba_output_stride_x Stride of the destination image in X dimension (in bytes) |
| 1219 | * @param[in] rgba_output_step_x rgba_output_stride_x * number of elements along X processed per workitem(in bytes) |
| 1220 | * @param[in] rgba_output_stride_y Stride of the destination image in Y dimension (in bytes) |
| 1221 | * @param[in] rgba_output_step_y rgba_output_stride_y * number of elements along Y processed per workitem(in bytes) |
| 1222 | * @param[in] rgba_output_offset_first_element_in_bytes The offset of the first element in the destination image |
| 1223 | */ |
| 1224 | __kernel void NV21_to_RGBA8888_bt709( |
| 1225 | IMAGE_DECLARATION(luma_input), |
| 1226 | IMAGE_DECLARATION(uv_input), |
| 1227 | IMAGE_DECLARATION(rgba_output)) |
| 1228 | { |
| 1229 | Image in_luma = CONVERT_TO_IMAGE_STRUCT(luma_input); |
| 1230 | Image in_uv = CONVERT_TO_IMAGE_STRUCT(uv_input); |
| 1231 | Image out_rgb = CONVERT_TO_IMAGE_STRUCT(rgba_output); |
| 1232 | |
| 1233 | // handle 8 pixels every time, two lines, each line for 4 pixels |
| 1234 | uchar4 luma_0 = vload4(0, in_luma.ptr); |
| 1235 | uchar4 luma_1 = vload4(0, in_luma.ptr + luma_input_stride_y); |
| 1236 | uchar4 cbcr = vload4(0, in_uv.ptr); |
| 1237 | char4 cr = (char4)(cbcr.s0, cbcr.s0, cbcr.s2, cbcr.s2) - (char4)(128); |
| 1238 | char4 cb = (char4)(cbcr.s1, cbcr.s1, cbcr.s3, cbcr.s3) - (char4)(128); |
| 1239 | |
| 1240 | float4 temp0 = (float4)(0.0000f) + (float4)(0.0000f) * convert_float4(cb) + (float4)(1.5748f) * convert_float4(cr); |
| 1241 | float4 temp1 = (float4)(0.0000f) - (float4)(0.1873f) * convert_float4(cb) - (float4)(0.4681f) * convert_float4(cr); |
| 1242 | float4 temp2 = (float4)(0.0000f) + (float4)(1.8556f) * convert_float4(cb) + (float4)(0.0000f) * convert_float4(cr); |
| 1243 | |
| 1244 | float4 f_r = convert_float4(luma_0) + temp0; |
| 1245 | float4 f_g = convert_float4(luma_0) + temp1; |
| 1246 | float4 f_b = convert_float4(luma_0) + temp2; |
| 1247 | |
Pablo Tello | 96fc1d6 | 2018-07-17 17:10:59 +0100 | [diff] [blame] | 1248 | uchar4 r_0 = convert_uchar4_sat_rtz(f_r); |
| 1249 | uchar4 g_0 = convert_uchar4_sat_rtz(f_g); |
| 1250 | uchar4 b_0 = convert_uchar4_sat_rtz(f_b); |
Anthony Barbier | 6ff3b19 | 2017-09-04 18:44:23 +0100 | [diff] [blame] | 1251 | |
| 1252 | uchar8 rgb_0 = (uchar8)(r_0.s0, g_0.s0, b_0.s0, 255, r_0.s1, g_0.s1, b_0.s1, 255); |
| 1253 | uchar8 rgb_1 = (uchar8)(r_0.s2, g_0.s2, b_0.s2, 255, r_0.s3, g_0.s3, b_0.s3, 255); |
| 1254 | vstore8(rgb_0, 0, out_rgb.ptr); |
| 1255 | vstore8(rgb_1, 0, out_rgb.ptr + 8); |
| 1256 | |
| 1257 | f_r = convert_float4(luma_1) + temp0; |
| 1258 | f_g = convert_float4(luma_1) + temp1; |
| 1259 | f_b = convert_float4(luma_1) + temp2; |
| 1260 | |
Pablo Tello | 96fc1d6 | 2018-07-17 17:10:59 +0100 | [diff] [blame] | 1261 | r_0 = convert_uchar4_sat_rtz(f_r); |
| 1262 | g_0 = convert_uchar4_sat_rtz(f_g); |
| 1263 | b_0 = convert_uchar4_sat_rtz(f_b); |
Anthony Barbier | 6ff3b19 | 2017-09-04 18:44:23 +0100 | [diff] [blame] | 1264 | |
| 1265 | rgb_0 = (uchar8)(r_0.s0, g_0.s0, b_0.s0, 255, r_0.s1, g_0.s1, b_0.s1, 255); |
| 1266 | rgb_1 = (uchar8)(r_0.s2, g_0.s2, b_0.s2, 255, r_0.s3, g_0.s3, b_0.s3, 255); |
| 1267 | vstore8(rgb_0, 0, out_rgb.ptr + rgba_output_stride_y); |
| 1268 | vstore8(rgb_1, 0, out_rgb.ptr + rgba_output_stride_y + 8); |
| 1269 | } |
| 1270 | |
| 1271 | /** Convert an NV21 image to YUV444 |
| 1272 | * |
| 1273 | * Global Workgroup Size [ DIV_CEIL(width, 16), height ] |
| 1274 | * No offset. |
| 1275 | * |
| 1276 | * @param[in] luma_input_ptr Pointer to the source luma channel. Supported Format: U8 |
| 1277 | * @param[in] luma_input_stride_x Stride of the luma image in X dimension (in bytes) |
| 1278 | * @param[in] luma_input_step_x luma_input_stride_x * number of elements along X processed per workitem(in bytes) |
| 1279 | * @param[in] luma_input_stride_y Stride of the source luma channel in Y dimension (in bytes) |
| 1280 | * @param[in] luma_input_step_y luma_input_stride_y * number of elements along Y processed per workitem(in bytes) |
| 1281 | * @param[in] luma_input_offset_first_element_in_bytes The offset of the first element in the source image |
| 1282 | * @param[in] uv_input_ptr Pointer to the source uv channel. Supported Format: U8 |
| 1283 | * @param[in] uv_input_stride_x Stride of the source image uv channel in X dimension (in bytes) |
| 1284 | * @param[in] uv_input_step_x uv_input_stride_x * number of elements along X processed per workitem(in bytes) |
| 1285 | * @param[in] uv_input_stride_y Stride of the source image in Y dimension (in bytes) |
| 1286 | * @param[in] uv_input_step_y uv_input_stride_y * number of elements along Y processed per workitem(in bytes) |
| 1287 | * @param[in] uv_input_offset_first_element_in_bytes The offset of the first element in the source image |
| 1288 | * @param[out] luma_output_ptr Pointer to the destination luma channel. Supported Format: U8 |
| 1289 | * @param[in] luma_output_stride_x Stride of the destination luma channel in X dimension (in bytes) |
| 1290 | * @param[in] luma_output_step_x luma_output_stride_x * number of elements along X processed per workitem(in bytes) |
| 1291 | * @param[in] luma_output_stride_y Stride of the destination image luma channel in Y dimension (in bytes) |
| 1292 | * @param[in] luma_output_step_y luma_output_stride_y * number of elements along Y processed per workitem(in bytes) |
| 1293 | * @param[in] luma_output_offset_first_element_in_bytes The offset of the first element in the destination luma channel |
| 1294 | * @param[out] u_output_ptr Pointer to the destination U channel. Supported Format: U8 |
| 1295 | * @param[in] u_output_stride_x Stride of the destination U channel in X dimension (in bytes) |
| 1296 | * @param[in] u_output_step_x u_output_stride_x * number of elements along X processed per workitem(in bytes) |
| 1297 | * @param[in] u_output_stride_y Stride of the destination image U channel in Y dimension (in bytes) |
| 1298 | * @param[in] u_output_step_y u_output_stride_y * number of elements along Y processed per workitem(in bytes) |
| 1299 | * @param[in] u_output_offset_first_element_in_bytes The offset of the first element in the destination U channel |
| 1300 | * @param[out] v_output_ptr Pointer to the destination V channel. Supported Format: U8 |
| 1301 | * @param[in] v_output_stride_x Stride of the destination V channel in X dimension (in bytes) |
| 1302 | * @param[in] v_output_step_x v_output_stride_x * number of elements along X processed per workitem(in bytes) |
| 1303 | * @param[in] v_output_stride_y Stride of the destination V channel in Y dimension (in bytes) |
| 1304 | * @param[in] v_output_step_y v_output_stride_y * number of elements along Y processed per workitem(in bytes) |
| 1305 | * @param[in] v_output_offset_first_element_in_bytes The offset of the first element in the destination V channel |
| 1306 | */ |
| 1307 | __kernel void NV21_to_YUV444_bt709( |
| 1308 | IMAGE_DECLARATION(luma_input), |
| 1309 | IMAGE_DECLARATION(uv_input), |
| 1310 | IMAGE_DECLARATION(luma_output), |
| 1311 | IMAGE_DECLARATION(u_output), |
| 1312 | IMAGE_DECLARATION(v_output)) |
| 1313 | { |
| 1314 | Image in_y = CONVERT_TO_IMAGE_STRUCT(luma_input); |
| 1315 | Image in_uv = CONVERT_TO_IMAGE_STRUCT(uv_input); |
| 1316 | Image out_y = CONVERT_TO_IMAGE_STRUCT(luma_output); |
| 1317 | Image out_u = CONVERT_TO_IMAGE_STRUCT(u_output); |
| 1318 | Image out_v = CONVERT_TO_IMAGE_STRUCT(v_output); |
| 1319 | |
| 1320 | // handle 32 pixels every time, two lines, each line for 16 pixels |
| 1321 | uchar16 luma_0 = vload16(0, in_y.ptr); |
| 1322 | uchar16 luma_1 = vload16(0, in_y.ptr + luma_input_stride_y); |
| 1323 | uchar16 cbcr = vload16(0, in_uv.ptr); |
| 1324 | uchar16 cr = (uchar16)(cbcr.s0, cbcr.s0, cbcr.s2, cbcr.s2, cbcr.s4, cbcr.s4, cbcr.s6, cbcr.s6, cbcr.s8, cbcr.s8, |
| 1325 | cbcr.sa, cbcr.sa, cbcr.sc, cbcr.sc, cbcr.se, cbcr.se); |
| 1326 | uchar16 cb = (uchar16)(cbcr.s1, cbcr.s1, cbcr.s3, cbcr.s3, cbcr.s5, cbcr.s5, cbcr.s7, cbcr.s7, cbcr.s9, cbcr.s9, |
| 1327 | cbcr.sb, cbcr.sb, cbcr.sd, cbcr.sd, cbcr.sf, cbcr.sf); |
| 1328 | |
| 1329 | vstore16(luma_0, 0, out_y.ptr); |
| 1330 | vstore16(luma_1, 0, out_y.ptr + luma_output_stride_y); |
| 1331 | vstore16(cb, 0, out_u.ptr); |
| 1332 | vstore16(cb, 0, out_u.ptr + u_output_stride_y); |
| 1333 | vstore16(cr, 0, out_v.ptr); |
| 1334 | vstore16(cr, 0, out_v.ptr + v_output_stride_y); |
| 1335 | } |
| 1336 | |
| 1337 | /** Convert an NV21 image to IYUV |
| 1338 | * |
| 1339 | * Global Workgroup Size [ DIV_CEIL(width, 16), height ] |
| 1340 | * No offset. |
| 1341 | * |
| 1342 | * @param[in] luma_input_ptr Pointer to the source luma channel. Supported Format: U8 |
| 1343 | * @param[in] luma_input_stride_x Stride of the luma image in X dimension (in bytes) |
| 1344 | * @param[in] luma_input_step_x luma_input_stride_x * number of elements along X processed per workitem(in bytes) |
| 1345 | * @param[in] luma_input_stride_y Stride of the source luma channel in Y dimension (in bytes) |
| 1346 | * @param[in] luma_input_step_y luma_input_stride_y * number of elements along Y processed per workitem(in bytes) |
| 1347 | * @param[in] luma_input_offset_first_element_in_bytes The offset of the first element in the source image |
| 1348 | * @param[in] uv_input_ptr Pointer to the source uv channel. Supported Format: U8 |
| 1349 | * @param[in] uv_input_stride_x Stride of the source image uv channel in X dimension (in bytes) |
| 1350 | * @param[in] uv_input_step_x uv_input_stride_x * number of elements along X processed per workitem(in bytes) |
| 1351 | * @param[in] uv_input_stride_y Stride of the source image in Y dimension (in bytes) |
| 1352 | * @param[in] uv_input_step_y uv_input_stride_y * number of elements along Y processed per workitem(in bytes) |
| 1353 | * @param[in] uv_input_offset_first_element_in_bytes The offset of the first element in the source image |
| 1354 | * @param[out] luma_output_ptr Pointer to the destination luma channel. Supported Format: U8 |
| 1355 | * @param[in] luma_output_stride_x Stride of the destination luma channel in X dimension (in bytes) |
| 1356 | * @param[in] luma_output_step_x luma_output_stride_x * number of elements along X processed per workitem(in bytes) |
| 1357 | * @param[in] luma_output_stride_y Stride of the destination image luma channel in Y dimension (in bytes) |
| 1358 | * @param[in] luma_output_step_y luma_output_stride_y * number of elements along Y processed per workitem(in bytes) |
| 1359 | * @param[in] luma_output_offset_first_element_in_bytes The offset of the first element in the destination luma channel |
| 1360 | * @param[out] u_output_ptr Pointer to the destination U channel. Supported Format: U8 |
| 1361 | * @param[in] u_output_stride_x Stride of the destination U channel in X dimension (in bytes) |
| 1362 | * @param[in] u_output_step_x u_output_stride_x * number of elements along X processed per workitem(in bytes) |
| 1363 | * @param[in] u_output_stride_y Stride of the destination image U channel in Y dimension (in bytes) |
| 1364 | * @param[in] u_output_step_y u_output_stride_y * number of elements along Y processed per workitem(in bytes) |
| 1365 | * @param[in] u_output_offset_first_element_in_bytes The offset of the first element in the destination U channel |
| 1366 | * @param[out] v_output_ptr Pointer to the destination V channel. Supported Format: U8 |
| 1367 | * @param[in] v_output_stride_x Stride of the destination V channel in X dimension (in bytes) |
| 1368 | * @param[in] v_output_step_x v_output_stride_x * number of elements along X processed per workitem(in bytes) |
| 1369 | * @param[in] v_output_stride_y Stride of the destination V channel in Y dimension (in bytes) |
| 1370 | * @param[in] v_output_step_y v_output_stride_y * number of elements along Y processed per workitem(in bytes) |
| 1371 | * @param[in] v_output_offset_first_element_in_bytes The offset of the first element in the destination V channel |
| 1372 | */ |
| 1373 | __kernel void NV21_to_IYUV_bt709( |
| 1374 | IMAGE_DECLARATION(luma_input), |
| 1375 | IMAGE_DECLARATION(uv_input), |
| 1376 | IMAGE_DECLARATION(luma_output), |
| 1377 | IMAGE_DECLARATION(u_output), |
| 1378 | IMAGE_DECLARATION(v_output)) |
| 1379 | { |
| 1380 | Image in_y = CONVERT_TO_IMAGE_STRUCT(luma_input); |
| 1381 | Image in_uv = CONVERT_TO_IMAGE_STRUCT(uv_input); |
| 1382 | Image out_y = CONVERT_TO_IMAGE_STRUCT(luma_output); |
| 1383 | Image out_u = CONVERT_TO_IMAGE_STRUCT(u_output); |
| 1384 | Image out_v = CONVERT_TO_IMAGE_STRUCT(v_output); |
| 1385 | |
| 1386 | uchar16 luma_0 = vload16(0, in_y.ptr); |
| 1387 | uchar16 luma_1 = vload16(0, in_y.ptr + luma_input_stride_y); |
| 1388 | uchar16 cbcr = vload16(0, in_uv.ptr); |
| 1389 | uchar8 cr = (uchar8)(cbcr.s0, cbcr.s2, cbcr.s4, cbcr.s6, cbcr.s8, cbcr.sa, cbcr.sc, cbcr.se); |
| 1390 | uchar8 cb = (uchar8)(cbcr.s1, cbcr.s3, cbcr.s5, cbcr.s7, cbcr.s9, cbcr.sb, cbcr.sd, cbcr.sf); |
| 1391 | |
| 1392 | vstore16(luma_0, 0, out_y.ptr); |
| 1393 | vstore16(luma_1, 0, out_y.ptr + luma_output_stride_y); |
| 1394 | vstore8(cb, 0, out_u.ptr); |
| 1395 | vstore8(cr, 0, out_v.ptr); |
| 1396 | } |
| 1397 | |
| 1398 | /** Convert a UYVY image to IYUV using BT709 color space |
| 1399 | * |
| 1400 | * Global Workgroup Size [ DIV_CEIL(width, 8), height ] |
| 1401 | * No offset. |
| 1402 | * |
| 1403 | * @param[in] uyvy_input_ptr Pointer to the source image. Supported Format: U8 |
| 1404 | * @param[in] uyvy_input_stride_x Stride of the source image in X dimension (in bytes) |
| 1405 | * @param[in] uyvy_input_step_x uyvy_input_stride_x * number of elements along X processed per workitem(in bytes) |
| 1406 | * @param[in] uyvy_input_stride_y Stride of the source image in Y dimension (in bytes) |
| 1407 | * @param[in] uyvy_input_step_y uyvy_input_stride_y * number of elements along Y processed per workitem(in bytes) |
| 1408 | * @param[in] uyvy_input_offset_first_element_in_bytes The offset of the first element in the source image |
| 1409 | * @param[out] luma_output_ptr Pointer to the destination luma channel. Supported Format: U8 |
| 1410 | * @param[in] luma_output_stride_x Stride of the destination luma channel in X dimension (in bytes) |
| 1411 | * @param[in] luma_output_step_x luma_output_stride_x * number of elements along X processed per workitem(in bytes) |
| 1412 | * @param[in] luma_output_stride_y Stride of the destination image luma channel in Y dimension (in bytes) |
| 1413 | * @param[in] luma_output_step_y luma_output_stride_y * number of elements along Y processed per workitem(in bytes) |
| 1414 | * @param[in] luma_output_offset_first_element_in_bytes The offset of the first element in the destination luma channel |
| 1415 | * @param[out] u_output_ptr Pointer to the destination U channel. Supported Format: U8 |
| 1416 | * @param[in] u_output_stride_x Stride of the destination U channel in X dimension (in bytes) |
| 1417 | * @param[in] u_output_step_x u_output_stride_x * number of elements along X processed per workitem(in bytes) |
| 1418 | * @param[in] u_output_stride_y Stride of the destination image U channel in Y dimension (in bytes) |
| 1419 | * @param[in] u_output_step_y u_output_stride_y * number of elements along Y processed per workitem(in bytes) |
| 1420 | * @param[in] u_output_offset_first_element_in_bytes The offset of the first element in the destination U channel |
| 1421 | * @param[out] v_output_ptr Pointer to the destination V channel. Supported Format: U8 |
| 1422 | * @param[in] v_output_stride_x Stride of the destination V channel in X dimension (in bytes) |
| 1423 | * @param[in] v_output_step_x v_output_stride_x * number of elements along X processed per workitem(in bytes) |
| 1424 | * @param[in] v_output_stride_y Stride of the destination V channel in Y dimension (in bytes) |
| 1425 | * @param[in] v_output_step_y v_output_stride_y * number of elements along Y processed per workitem(in bytes) |
| 1426 | * @param[in] v_output_offset_first_element_in_bytes The offset of the first element in the destination V channel |
| 1427 | * |
| 1428 | */ |
| 1429 | __kernel void UYVY422_to_IYUV_bt709( |
| 1430 | IMAGE_DECLARATION(uyvy_input), |
| 1431 | IMAGE_DECLARATION(luma_output), |
| 1432 | IMAGE_DECLARATION(u_output), |
| 1433 | IMAGE_DECLARATION(v_output)) |
| 1434 | { |
| 1435 | Image in_uyvy = CONVERT_TO_IMAGE_STRUCT(uyvy_input); |
| 1436 | Image out_y = CONVERT_TO_IMAGE_STRUCT(luma_output); |
| 1437 | Image out_u = CONVERT_TO_IMAGE_STRUCT(u_output); |
| 1438 | Image out_v = CONVERT_TO_IMAGE_STRUCT(v_output); |
| 1439 | |
| 1440 | // handle 16 pixels every time, each line 8 pixels |
| 1441 | uchar16 uyvy = vload16(0, in_uyvy.ptr); |
| 1442 | uchar8 luma = (uchar8)(uyvy.s1, uyvy.s3, uyvy.s5, uyvy.s7, uyvy.s9, uyvy.sb, uyvy.sd, uyvy.sf); |
| 1443 | ushort4 cb_0 = (ushort4)(uyvy.s0, uyvy.s4, uyvy.s8, uyvy.sc); |
| 1444 | ushort4 cr_0 = (ushort4)(uyvy.s2, uyvy.s6, uyvy.sa, uyvy.se); |
| 1445 | vstore8(luma, 0, out_y.ptr); |
| 1446 | |
| 1447 | uyvy = vload16(0, in_uyvy.ptr + uyvy_input_stride_y); |
| 1448 | luma = (uchar8)(uyvy.s1, uyvy.s3, uyvy.s5, uyvy.s7, uyvy.s9, uyvy.sb, uyvy.sd, uyvy.sf); |
| 1449 | ushort4 cb_1 = (ushort4)(uyvy.s0, uyvy.s4, uyvy.s8, uyvy.sc); |
| 1450 | ushort4 cr_1 = (ushort4)(uyvy.s2, uyvy.s6, uyvy.sa, uyvy.se); |
| 1451 | vstore8(luma, 0, out_y.ptr + luma_output_stride_y); |
| 1452 | |
| 1453 | uchar4 cb = convert_uchar4((cb_0 + cb_1) / (ushort4)(2)); |
| 1454 | uchar4 cr = convert_uchar4((cr_0 + cr_1) / (ushort4)(2)); |
| 1455 | vstore4(cb, 0, out_u.ptr); |
| 1456 | vstore4(cr, 0, out_v.ptr); |
| 1457 | } |
| 1458 | |
| 1459 | /** Convert a YUYV image to IYUV using BT709 color space |
| 1460 | * |
| 1461 | * Global Workgroup Size [ DIV_CEIL(width, 8), height ] |
| 1462 | * No offset. |
| 1463 | * |
| 1464 | * @param[in] yuyv_input_ptr Pointer to the source image. Supported Format: U8 |
| 1465 | * @param[in] yuyv_input_stride_x Stride of the source image in X dimension (in bytes) |
| 1466 | * @param[in] yuyv_input_step_x yuyv_input_stride_x * number of elements along X processed per workitem(in bytes) |
| 1467 | * @param[in] yuyv_input_stride_y Stride of the source image in Y dimension (in bytes) |
| 1468 | * @param[in] yuyv_input_step_y yuyv_input_stride_y * number of elements along Y processed per workitem(in bytes) |
| 1469 | * @param[in] yuyv_input_offset_first_element_in_bytes The offset of the first element in the source image |
| 1470 | * @param[out] luma_output_ptr Pointer to the destination luma channel. Supported Format: U8 |
| 1471 | * @param[in] luma_output_stride_x Stride of the destination luma channel in X dimension (in bytes) |
| 1472 | * @param[in] luma_output_step_x luma_output_stride_x * number of elements along X processed per workitem(in bytes) |
| 1473 | * @param[in] luma_output_stride_y Stride of the destination image luma channel in Y dimension (in bytes) |
| 1474 | * @param[in] luma_output_step_y luma_output_stride_y * number of elements along Y processed per workitem(in bytes) |
| 1475 | * @param[in] luma_output_offset_first_element_in_bytes The offset of the first element in the destination luma channel |
| 1476 | * @param[out] u_output_ptr Pointer to the destination U channel. Supported Format: U8 |
| 1477 | * @param[in] u_output_stride_x Stride of the destination U channel in X dimension (in bytes) |
| 1478 | * @param[in] u_output_step_x u_output_stride_x * number of elements along X processed per workitem(in bytes) |
| 1479 | * @param[in] u_output_stride_y Stride of the destination image U channel in Y dimension (in bytes) |
| 1480 | * @param[in] u_output_step_y u_output_stride_y * number of elements along Y processed per workitem(in bytes) |
| 1481 | * @param[in] u_output_offset_first_element_in_bytes The offset of the first element in the destination U channel |
| 1482 | * @param[out] v_output_ptr Pointer to the destination V channel. Supported Format: U8 |
| 1483 | * @param[in] v_output_stride_x Stride of the destination V channel in X dimension (in bytes) |
| 1484 | * @param[in] v_output_step_x v_output_stride_x * number of elements along X processed per workitem(in bytes) |
| 1485 | * @param[in] v_output_stride_y Stride of the destination V channel in Y dimension (in bytes) |
| 1486 | * @param[in] v_output_step_y v_output_stride_y * number of elements along Y processed per workitem(in bytes) |
| 1487 | * @param[in] v_output_offset_first_element_in_bytes The offset of the first element in the destination V channel |
| 1488 | * |
| 1489 | */ |
| 1490 | __kernel void YUYV422_to_IYUV_bt709( |
| 1491 | IMAGE_DECLARATION(yuyv_input), |
| 1492 | IMAGE_DECLARATION(luma_output), |
| 1493 | IMAGE_DECLARATION(u_output), |
| 1494 | IMAGE_DECLARATION(v_output)) |
| 1495 | { |
| 1496 | Image in_yuyv = CONVERT_TO_IMAGE_STRUCT(yuyv_input); |
| 1497 | Image out_y = CONVERT_TO_IMAGE_STRUCT(luma_output); |
| 1498 | Image out_u = CONVERT_TO_IMAGE_STRUCT(u_output); |
| 1499 | Image out_v = CONVERT_TO_IMAGE_STRUCT(v_output); |
| 1500 | |
| 1501 | // handle 16 pixels every time, each line 8 pixels |
| 1502 | uchar16 yuyv = vload16(0, in_yuyv.ptr); |
| 1503 | uchar8 luma = (uchar8)(yuyv.s0, yuyv.s2, yuyv.s4, yuyv.s6, yuyv.s8, yuyv.sa, yuyv.sc, yuyv.se); |
| 1504 | ushort4 cb_0 = (ushort4)(yuyv.s1, yuyv.s5, yuyv.s9, yuyv.sd); |
| 1505 | ushort4 cr_0 = (ushort4)(yuyv.s3, yuyv.s7, yuyv.sb, yuyv.sf); |
| 1506 | vstore8(luma, 0, out_y.ptr); |
| 1507 | |
| 1508 | yuyv = vload16(0, in_yuyv.ptr + yuyv_input_stride_y); |
| 1509 | luma = (uchar8)(yuyv.s0, yuyv.s2, yuyv.s4, yuyv.s6, yuyv.s8, yuyv.sa, yuyv.sc, yuyv.se); |
| 1510 | ushort4 cb_1 = (ushort4)(yuyv.s1, yuyv.s5, yuyv.s9, yuyv.sd); |
| 1511 | ushort4 cr_1 = (ushort4)(yuyv.s3, yuyv.s7, yuyv.sb, yuyv.sf); |
| 1512 | vstore8(luma, 0, out_y.ptr + luma_output_stride_y); |
| 1513 | |
| 1514 | uchar4 cb = convert_uchar4((cb_0 + cb_1) / (ushort4)(2)); |
| 1515 | uchar4 cr = convert_uchar4((cr_0 + cr_1) / (ushort4)(2)); |
| 1516 | vstore4(cb, 0, out_u.ptr); |
| 1517 | vstore4(cr, 0, out_v.ptr); |
| 1518 | } |
| 1519 | |
| 1520 | /** Convert an IYUV image to RGB888 |
| 1521 | * |
| 1522 | * Global Workgroup Size [ DIV_CEIL(width, 4), height ] |
| 1523 | * No offset. |
| 1524 | * |
| 1525 | * @param[in] luma_input_ptr Pointer to the source luma channel. Supported Format: U8 |
| 1526 | * @param[in] luma_input_stride_x Stride of the luma image in X dimension (in bytes) |
| 1527 | * @param[in] luma_input_step_x luma_input_stride_x * number of elements along X processed per workitem(in bytes) |
| 1528 | * @param[in] luma_input_stride_y Stride of the source luma channel in Y dimension (in bytes) |
| 1529 | * @param[in] luma_input_step_y luma_input_stride_y * number of elements along Y processed per workitem(in bytes) |
| 1530 | * @param[in] luma_input_offset_first_element_in_bytes The offset of the first element in the source image |
| 1531 | * @param[in] u_input_ptr Pointer to the source U channel. Supported Format: U8 |
| 1532 | * @param[in] u_input_stride_x Stride of the source image U channel in X dimension (in bytes) |
| 1533 | * @param[in] u_input_step_x u_input_stride_x * number of elements along X processed per workitem(in bytes) |
| 1534 | * @param[in] u_input_stride_y Stride of the source image in Y dimension (in bytes) |
| 1535 | * @param[in] u_input_step_y u_input_stride_y * number of elements along Y processed per workitem(in bytes) |
| 1536 | * @param[in] u_input_offset_first_element_in_bytes The offset of the first element in the source U channel |
| 1537 | * @param[in] v_input_ptr Pointer to the source V channel. Supported Format: U8 |
| 1538 | * @param[in] v_input_stride_x Stride of the source image V channel in X dimension (in bytes) |
| 1539 | * @param[in] v_input_step_x v_input_stride_x * number of elements along X processed per workitem(in bytes) |
| 1540 | * @param[in] v_input_stride_y Stride of the source image V channel in Y dimension (in bytes) |
| 1541 | * @param[in] v_input_step_y v_input_stride_y * number of elements along Y processed per workitem(in bytes) |
| 1542 | * @param[in] v_input_offset_first_element_in_bytes The offset of the first element in the source image V channel |
| 1543 | * @param[out] rgb_output_ptr Pointer to the destination image. Supported Format: U8 |
| 1544 | * @param[in] rgb_output_stride_x Stride of the destination image in X dimension (in bytes) |
| 1545 | * @param[in] rgb_output_step_x rgb_output_stride_x * number of elements along X processed per workitem(in bytes) |
| 1546 | * @param[in] rgb_output_stride_y Stride of the destination image in Y dimension (in bytes) |
| 1547 | * @param[in] rgb_output_step_y rgb_output_stride_y * number of elements along Y processed per workitem(in bytes) |
| 1548 | * @param[in] rgb_output_offset_first_element_in_bytes The offset of the first element in the destination image |
| 1549 | */ |
| 1550 | __kernel void IYUV_to_RGB888_bt709( |
| 1551 | IMAGE_DECLARATION(luma_input), |
| 1552 | IMAGE_DECLARATION(u_input), |
| 1553 | IMAGE_DECLARATION(v_input), |
| 1554 | IMAGE_DECLARATION(rgb_output)) |
| 1555 | { |
| 1556 | Image in_y = CONVERT_TO_IMAGE_STRUCT(luma_input); |
| 1557 | Image in_u = CONVERT_TO_IMAGE_STRUCT(u_input); |
| 1558 | Image in_v = CONVERT_TO_IMAGE_STRUCT(v_input); |
| 1559 | Image out_rgb = CONVERT_TO_IMAGE_STRUCT(rgb_output); |
| 1560 | |
| 1561 | // handle 8 pixels every time, two lines, each line for 4 pixels |
| 1562 | uchar4 luma_0 = vload4(0, in_y.ptr); |
| 1563 | uchar4 luma_1 = vload4(0, in_y.ptr + luma_input_stride_y); |
| 1564 | uchar4 cbcr = (uchar4)(vload2(0, in_u.ptr), vload2(0, in_v.ptr)); |
| 1565 | char4 cb = (char4)(cbcr.s0, cbcr.s0, cbcr.s1, cbcr.s1) - (char4)(128); |
| 1566 | char4 cr = (char4)(cbcr.s2, cbcr.s2, cbcr.s3, cbcr.s3) - (char4)(128); |
| 1567 | |
| 1568 | float4 temp0 = (float4)(0.0000f) + (float4)(0.0000f) * convert_float4(cb) + (float4)(1.5748f) * convert_float4(cr); |
| 1569 | float4 temp1 = (float4)(0.0000f) - (float4)(0.1873f) * convert_float4(cb) - (float4)(0.4681f) * convert_float4(cr); |
| 1570 | float4 temp2 = (float4)(0.0000f) + (float4)(1.8556f) * convert_float4(cb) + (float4)(0.0000f) * convert_float4(cr); |
| 1571 | |
| 1572 | float4 f_r = convert_float4(luma_0) + temp0; |
| 1573 | float4 f_g = convert_float4(luma_0) + temp1; |
| 1574 | float4 f_b = convert_float4(luma_0) + temp2; |
| 1575 | |
Pablo Tello | 96fc1d6 | 2018-07-17 17:10:59 +0100 | [diff] [blame] | 1576 | uchar4 r_0 = convert_uchar4_sat_rtz(f_r); |
| 1577 | uchar4 g_0 = convert_uchar4_sat_rtz(f_g); |
| 1578 | uchar4 b_0 = convert_uchar4_sat_rtz(f_b); |
Anthony Barbier | 6ff3b19 | 2017-09-04 18:44:23 +0100 | [diff] [blame] | 1579 | |
| 1580 | uchar8 rgb_0 = (uchar8)(r_0.s0, g_0.s0, b_0.s0, r_0.s1, g_0.s1, b_0.s1, r_0.s2, g_0.s2); |
| 1581 | uchar4 rgb_1 = (uchar4)(b_0.s2, r_0.s3, g_0.s3, b_0.s3); |
| 1582 | vstore8(rgb_0, 0, out_rgb.ptr); |
| 1583 | vstore4(rgb_1, 0, out_rgb.ptr + 8); |
| 1584 | |
| 1585 | f_r = convert_float4(luma_1) + temp0; |
| 1586 | f_g = convert_float4(luma_1) + temp1; |
| 1587 | f_b = convert_float4(luma_1) + temp2; |
| 1588 | |
Pablo Tello | 96fc1d6 | 2018-07-17 17:10:59 +0100 | [diff] [blame] | 1589 | r_0 = convert_uchar4_sat_rtz(f_r); |
| 1590 | g_0 = convert_uchar4_sat_rtz(f_g); |
| 1591 | b_0 = convert_uchar4_sat_rtz(f_b); |
Anthony Barbier | 6ff3b19 | 2017-09-04 18:44:23 +0100 | [diff] [blame] | 1592 | |
| 1593 | rgb_0 = (uchar8)(r_0.s0, g_0.s0, b_0.s0, r_0.s1, g_0.s1, b_0.s1, r_0.s2, g_0.s2); |
| 1594 | rgb_1 = (uchar4)(b_0.s2, r_0.s3, g_0.s3, b_0.s3); |
| 1595 | vstore8(rgb_0, 0, out_rgb.ptr + rgb_output_stride_y); |
| 1596 | vstore4(rgb_1, 0, out_rgb.ptr + rgb_output_stride_y + 8); |
| 1597 | } |
| 1598 | |
| 1599 | /** Convert an IYUV image to RGB8888 |
| 1600 | * |
| 1601 | * Global Workgroup Size [ DIV_CEIL(width, 4), height ] |
| 1602 | * No offset. |
| 1603 | * |
| 1604 | * @param[in] luma_input_ptr Pointer to the source luma channel. Supported Format: U8 |
| 1605 | * @param[in] luma_input_stride_x Stride of the luma image in X dimension (in bytes) |
| 1606 | * @param[in] luma_input_step_x luma_input_stride_x * number of elements along X processed per workitem(in bytes) |
| 1607 | * @param[in] luma_input_stride_y Stride of the source luma channel in Y dimension (in bytes) |
| 1608 | * @param[in] luma_input_step_y luma_input_stride_y * number of elements along Y processed per workitem(in bytes) |
| 1609 | * @param[in] luma_input_offset_first_element_in_bytes The offset of the first element in the source image |
| 1610 | * @param[in] u_input_ptr Pointer to the source U channel. Supported Format: U8 |
| 1611 | * @param[in] u_input_stride_x Stride of the source image U channel in X dimension (in bytes) |
| 1612 | * @param[in] u_input_step_x u_input_stride_x * number of elements along X processed per workitem(in bytes) |
| 1613 | * @param[in] u_input_stride_y Stride of the source image in Y dimension (in bytes) |
| 1614 | * @param[in] u_input_step_y u_input_stride_y * number of elements along Y processed per workitem(in bytes) |
| 1615 | * @param[in] u_input_offset_first_element_in_bytes The offset of the first element in the source U channel |
| 1616 | * @param[in] v_input_ptr Pointer to the source V channel. Supported Format: U8 |
| 1617 | * @param[in] v_input_stride_x Stride of the source image V channel in X dimension (in bytes) |
| 1618 | * @param[in] v_input_step_x v_input_stride_x * number of elements along X processed per workitem(in bytes) |
| 1619 | * @param[in] v_input_stride_y Stride of the source image V channel in Y dimension (in bytes) |
| 1620 | * @param[in] v_input_step_y v_input_stride_y * number of elements along Y processed per workitem(in bytes) |
| 1621 | * @param[in] v_input_offset_first_element_in_bytes The offset of the first element in the source image V channel |
| 1622 | * @param[out] rgba_output_ptr Pointer to the destination image. Supported Format: U8 |
| 1623 | * @param[in] rgba_output_stride_x Stride of the destination image in X dimension (in bytes) |
| 1624 | * @param[in] rgba_output_step_x rgba_output_stride_x * number of elements along X processed per workitem(in bytes) |
| 1625 | * @param[in] rgba_output_stride_y Stride of the destination image in Y dimension (in bytes) |
| 1626 | * @param[in] rgba_output_step_y rgba_output_stride_y * number of elements along Y processed per workitem(in bytes) |
| 1627 | * @param[in] rgba_output_offset_first_element_in_bytes The offset of the first element in the destination image |
| 1628 | */ |
| 1629 | __kernel void IYUV_to_RGBA8888_bt709( |
| 1630 | IMAGE_DECLARATION(luma_input), |
| 1631 | IMAGE_DECLARATION(u_input), |
| 1632 | IMAGE_DECLARATION(v_input), |
| 1633 | IMAGE_DECLARATION(rgba_output)) |
| 1634 | { |
| 1635 | Image in_y = CONVERT_TO_IMAGE_STRUCT(luma_input); |
| 1636 | Image in_u = CONVERT_TO_IMAGE_STRUCT(u_input); |
| 1637 | Image in_v = CONVERT_TO_IMAGE_STRUCT(v_input); |
| 1638 | Image out_rgb = CONVERT_TO_IMAGE_STRUCT(rgba_output); |
| 1639 | |
| 1640 | // handle 8 pixels every time, two lines, each line for 4 pixels |
| 1641 | uchar4 luma_0 = vload4(0, in_y.ptr); |
| 1642 | uchar4 luma_1 = vload4(0, in_y.ptr + luma_input_stride_y); |
| 1643 | uchar4 cbcr = (uchar4)(vload2(0, in_u.ptr), vload2(0, in_v.ptr)); |
| 1644 | char4 cb = (char4)(cbcr.s0, cbcr.s0, cbcr.s1, cbcr.s1) - (char4)(128); |
| 1645 | char4 cr = (char4)(cbcr.s2, cbcr.s2, cbcr.s3, cbcr.s3) - (char4)(128); |
| 1646 | |
| 1647 | float4 temp0 = (float4)(0.0000f) + (float4)(0.0000f) * convert_float4(cb) + (float4)(1.5748f) * convert_float4(cr); |
| 1648 | float4 temp1 = (float4)(0.0000f) - (float4)(0.1873f) * convert_float4(cb) - (float4)(0.4681f) * convert_float4(cr); |
| 1649 | float4 temp2 = (float4)(0.0000f) + (float4)(1.8556f) * convert_float4(cb) + (float4)(0.0000f) * convert_float4(cr); |
| 1650 | |
| 1651 | float4 f_r = convert_float4(luma_0) + temp0; |
| 1652 | float4 f_g = convert_float4(luma_0) + temp1; |
| 1653 | float4 f_b = convert_float4(luma_0) + temp2; |
| 1654 | |
Pablo Tello | 96fc1d6 | 2018-07-17 17:10:59 +0100 | [diff] [blame] | 1655 | uchar4 r_0 = convert_uchar4_sat_rtz(f_r); |
| 1656 | uchar4 g_0 = convert_uchar4_sat_rtz(f_g); |
| 1657 | uchar4 b_0 = convert_uchar4_sat_rtz(f_b); |
Anthony Barbier | 6ff3b19 | 2017-09-04 18:44:23 +0100 | [diff] [blame] | 1658 | |
| 1659 | uchar8 rgb_0 = (uchar8)(r_0.s0, g_0.s0, b_0.s0, 255, r_0.s1, g_0.s1, b_0.s1, 255); |
| 1660 | uchar8 rgb_1 = (uchar8)(r_0.s2, g_0.s2, b_0.s2, 255, r_0.s3, g_0.s3, b_0.s3, 255); |
| 1661 | vstore8(rgb_0, 0, out_rgb.ptr); |
| 1662 | vstore8(rgb_1, 0, out_rgb.ptr + 8); |
| 1663 | |
| 1664 | f_r = convert_float4(luma_1) + temp0; |
| 1665 | f_g = convert_float4(luma_1) + temp1; |
| 1666 | f_b = convert_float4(luma_1) + temp2; |
| 1667 | |
Pablo Tello | 96fc1d6 | 2018-07-17 17:10:59 +0100 | [diff] [blame] | 1668 | r_0 = convert_uchar4_sat_rtz(f_r); |
| 1669 | g_0 = convert_uchar4_sat_rtz(f_g); |
| 1670 | b_0 = convert_uchar4_sat_rtz(f_b); |
Anthony Barbier | 6ff3b19 | 2017-09-04 18:44:23 +0100 | [diff] [blame] | 1671 | |
| 1672 | rgb_0 = (uchar8)(r_0.s0, g_0.s0, b_0.s0, 255, r_0.s1, g_0.s1, b_0.s1, 255); |
| 1673 | rgb_1 = (uchar8)(r_0.s2, g_0.s2, b_0.s2, 255, r_0.s3, g_0.s3, b_0.s3, 255); |
| 1674 | vstore8(rgb_0, 0, out_rgb.ptr + rgba_output_stride_y); |
| 1675 | vstore8(rgb_1, 0, out_rgb.ptr + rgba_output_stride_y + 8); |
| 1676 | } |
| 1677 | |
| 1678 | /** Convert an IYUV image to YUV444 |
| 1679 | * |
| 1680 | * Global Workgroup Size [ DIV_CEIL(width, 16), height ] |
| 1681 | * No offset. |
| 1682 | * |
| 1683 | * @param[in] luma_input_ptr Pointer to the source luma channel. Supported Format: U8 |
| 1684 | * @param[in] luma_input_stride_x Stride of the luma image in X dimension (in bytes) |
| 1685 | * @param[in] luma_input_step_x luma_input_stride_x * number of elements along X processed per workitem(in bytes) |
| 1686 | * @param[in] luma_input_stride_y Stride of the source luma channel in Y dimension (in bytes) |
| 1687 | * @param[in] luma_input_step_y luma_input_stride_y * number of elements along Y processed per workitem(in bytes) |
| 1688 | * @param[in] luma_input_offset_first_element_in_bytes The offset of the first element in the source image |
| 1689 | * @param[in] u_input_ptr Pointer to the source U channel. Supported Format: U8 |
| 1690 | * @param[in] u_input_stride_x Stride of the source image U channel in X dimension (in bytes) |
| 1691 | * @param[in] u_input_step_x u_input_stride_x * number of elements along X processed per workitem(in bytes) |
| 1692 | * @param[in] u_input_stride_y Stride of the source image in Y dimension (in bytes) |
| 1693 | * @param[in] u_input_step_y u_input_stride_y * number of elements along Y processed per workitem(in bytes) |
| 1694 | * @param[in] u_input_offset_first_element_in_bytes The offset of the first element in the source U channel |
| 1695 | * @param[in] v_input_ptr Pointer to the source V channel. Supported Format: U8 |
| 1696 | * @param[in] v_input_stride_x Stride of the source image V channel in X dimension (in bytes) |
| 1697 | * @param[in] v_input_step_x v_input_stride_x * number of elements along X processed per workitem(in bytes) |
| 1698 | * @param[in] v_input_stride_y Stride of the source image V channel in Y dimension (in bytes) |
| 1699 | * @param[in] v_input_step_y v_input_stride_y * number of elements along Y processed per workitem(in bytes) |
| 1700 | * @param[in] v_input_offset_first_element_in_bytes The offset of the first element in the source image V channel |
| 1701 | * @param[out] luma_output_ptr Pointer to the destination luma channel. Supported Format: U8 |
| 1702 | * @param[in] luma_output_stride_x Stride of the destination luma channel in X dimension (in bytes) |
| 1703 | * @param[in] luma_output_step_x luma_output_stride_x * number of elements along X processed per workitem(in bytes) |
| 1704 | * @param[in] luma_output_stride_y Stride of the destination image luma channel in Y dimension (in bytes) |
| 1705 | * @param[in] luma_output_step_y luma_output_stride_y * number of elements along Y processed per workitem(in bytes) |
| 1706 | * @param[in] luma_output_offset_first_element_in_bytes The offset of the first element in the destination luma channel |
| 1707 | * @param[out] u_output_ptr Pointer to the destination U channel. Supported Format: U8 |
| 1708 | * @param[in] u_output_stride_x Stride of the destination U channel in X dimension (in bytes) |
| 1709 | * @param[in] u_output_step_x u_output_stride_x * number of elements along X processed per workitem(in bytes) |
| 1710 | * @param[in] u_output_stride_y Stride of the destination image U channel in Y dimension (in bytes) |
| 1711 | * @param[in] u_output_step_y u_output_stride_y * number of elements along Y processed per workitem(in bytes) |
| 1712 | * @param[in] u_output_offset_first_element_in_bytes The offset of the first element in the destination U channel |
| 1713 | * @param[out] v_output_ptr Pointer to the destination V channel. Supported Format: U8 |
| 1714 | * @param[in] v_output_stride_x Stride of the destination V channel in X dimension (in bytes) |
| 1715 | * @param[in] v_output_step_x v_output_stride_x * number of elements along X processed per workitem(in bytes) |
| 1716 | * @param[in] v_output_stride_y Stride of the destination V channel in Y dimension (in bytes) |
| 1717 | * @param[in] v_output_step_y v_output_stride_y * number of elements along Y processed per workitem(in bytes) |
| 1718 | * @param[in] v_output_offset_first_element_in_bytes The offset of the first element in the destination V channel |
| 1719 | * |
| 1720 | */ |
| 1721 | __kernel void IYUV_to_YUV444_bt709( |
| 1722 | IMAGE_DECLARATION(luma_input), |
| 1723 | IMAGE_DECLARATION(u_input), |
| 1724 | IMAGE_DECLARATION(v_input), |
| 1725 | IMAGE_DECLARATION(luma_output), |
| 1726 | IMAGE_DECLARATION(u_output), |
| 1727 | IMAGE_DECLARATION(v_output)) |
| 1728 | { |
| 1729 | Image in_y = CONVERT_TO_IMAGE_STRUCT(luma_input); |
| 1730 | Image in_u = CONVERT_TO_IMAGE_STRUCT(u_input); |
| 1731 | Image in_v = CONVERT_TO_IMAGE_STRUCT(v_input); |
| 1732 | Image out_y = CONVERT_TO_IMAGE_STRUCT(luma_output); |
| 1733 | Image out_u = CONVERT_TO_IMAGE_STRUCT(u_output); |
| 1734 | Image out_v = CONVERT_TO_IMAGE_STRUCT(v_output); |
| 1735 | |
| 1736 | // handle 32 pixels every time, two lines, each line for 16 pixels |
| 1737 | uchar16 luma_0 = vload16(0, in_y.ptr); |
| 1738 | uchar16 luma_1 = vload16(0, in_y.ptr + luma_input_stride_y); |
| 1739 | uchar8 cb_src = vload8(0, in_u.ptr); |
| 1740 | uchar8 cr_src = vload8(0, in_v.ptr); |
| 1741 | uchar16 cb = (uchar16)(cb_src.s0, cb_src.s0, cb_src.s1, cb_src.s1, cb_src.s2, cb_src.s2, cb_src.s3, cb_src.s3, |
| 1742 | cb_src.s4, cb_src.s4, cb_src.s5, cb_src.s5, cb_src.s6, cb_src.s6, cb_src.s7, cb_src.s7); |
| 1743 | uchar16 cr = (uchar16)(cr_src.s0, cr_src.s0, cr_src.s1, cr_src.s1, cr_src.s2, cr_src.s2, cr_src.s3, cr_src.s3, |
| 1744 | cr_src.s4, cr_src.s4, cr_src.s5, cr_src.s5, cr_src.s6, cr_src.s6, cr_src.s7, cr_src.s7); |
| 1745 | |
| 1746 | vstore16(luma_0, 0, out_y.ptr); |
| 1747 | vstore16(luma_1, 0, out_y.ptr + luma_output_stride_y); |
| 1748 | vstore16(cb, 0, out_u.ptr); |
| 1749 | vstore16(cb, 0, out_u.ptr + u_output_stride_y); |
| 1750 | vstore16(cr, 0, out_v.ptr); |
| 1751 | vstore16(cr, 0, out_v.ptr + v_output_stride_y); |
| 1752 | } |
| 1753 | |
| 1754 | /** Convert an IYUV image to NV12 |
| 1755 | * |
| 1756 | * Global Workgroup Size [ DIV_CEIL(width, 16), height ] |
| 1757 | * No offset. |
| 1758 | * |
| 1759 | * @param[in] luma_input_ptr Pointer to the source luma channel. Supported Format: U8 |
| 1760 | * @param[in] luma_input_stride_x Stride of the luma image in X dimension (in bytes) |
| 1761 | * @param[in] luma_input_step_x luma_input_stride_x * number of elements along X processed per workitem(in bytes) |
| 1762 | * @param[in] luma_input_stride_y Stride of the source luma channel in Y dimension (in bytes) |
| 1763 | * @param[in] luma_input_step_y luma_input_stride_y * number of elements along Y processed per workitem(in bytes) |
| 1764 | * @param[in] luma_input_offset_first_element_in_bytes The offset of the first element in the source image |
| 1765 | * @param[in] u_input_ptr Pointer to the source U channel. Supported Format: U8 |
| 1766 | * @param[in] u_input_stride_x Stride of the source image U channel in X dimension (in bytes) |
| 1767 | * @param[in] u_input_step_x u_input_stride_x * number of elements along X processed per workitem(in bytes) |
| 1768 | * @param[in] u_input_stride_y Stride of the source image in Y dimension (in bytes) |
| 1769 | * @param[in] u_input_step_y u_input_stride_y * number of elements along Y processed per workitem(in bytes) |
| 1770 | * @param[in] u_input_offset_first_element_in_bytes The offset of the first element in the source U channel |
| 1771 | * @param[in] v_input_ptr Pointer to the source V channel. Supported Format: U8 |
| 1772 | * @param[in] v_input_stride_x Stride of the source image V channel in X dimension (in bytes) |
| 1773 | * @param[in] v_input_step_x v_input_stride_x * number of elements along X processed per workitem(in bytes) |
| 1774 | * @param[in] v_input_stride_y Stride of the source image V channel in Y dimension (in bytes) |
| 1775 | * @param[in] v_input_step_y v_input_stride_y * number of elements along Y processed per workitem(in bytes) |
| 1776 | * @param[in] v_input_offset_first_element_in_bytes The offset of the first element in the source image V channel |
| 1777 | * @param[out] luma_output_ptr Pointer to the destination luma channel. Supported Format: U8 |
| 1778 | * @param[in] luma_output_stride_x Stride of the destination luma channel in X dimension (in bytes) |
| 1779 | * @param[in] luma_output_step_x luma_output_stride_x * number of elements along X processed per workitem(in bytes) |
| 1780 | * @param[in] luma_output_stride_y Stride of the destination image luma channel in Y dimension (in bytes) |
| 1781 | * @param[in] luma_output_step_y luma_output_stride_y * number of elements along Y processed per workitem(in bytes) |
| 1782 | * @param[in] luma_output_offset_first_element_in_bytes The offset of the first element in the destination luma channel |
| 1783 | * @param[out] uv_output_ptr Pointer to the destination UV channel. Supported Format: U8 |
| 1784 | * @param[in] uv_output_stride_x Stride of the destination UV channel in X dimension (in bytes) |
| 1785 | * @param[in] uv_output_step_x uv_output_stride_x * number of elements along X processed per workitem(in bytes) |
| 1786 | * @param[in] uv_output_stride_y Stride of the destination image U channel in Y dimension (in bytes) |
| 1787 | * @param[in] uv_output_step_y uv_output_stride_y * number of elements along Y processed per workitem(in bytes) |
| 1788 | * @param[in] uv_output_offset_first_element_in_bytes The offset of the first element in the destination UV channel |
| 1789 | * |
| 1790 | */ |
| 1791 | __kernel void IYUV_to_NV12_bt709( |
| 1792 | IMAGE_DECLARATION(luma_input), |
| 1793 | IMAGE_DECLARATION(u_input), |
| 1794 | IMAGE_DECLARATION(v_input), |
| 1795 | IMAGE_DECLARATION(luma_output), |
| 1796 | IMAGE_DECLARATION(uv_output)) |
| 1797 | { |
| 1798 | Image in_y = CONVERT_TO_IMAGE_STRUCT(luma_input); |
| 1799 | Image in_u = CONVERT_TO_IMAGE_STRUCT(u_input); |
| 1800 | Image in_v = CONVERT_TO_IMAGE_STRUCT(v_input); |
| 1801 | Image out_y = CONVERT_TO_IMAGE_STRUCT(luma_output); |
| 1802 | Image out_uv = CONVERT_TO_IMAGE_STRUCT(uv_output); |
| 1803 | |
| 1804 | // handle 32 pixels every time, two lines, each line for 16 pixels |
| 1805 | uchar16 luma_0 = vload16(0, in_y.ptr); |
| 1806 | uchar16 luma_1 = vload16(0, in_y.ptr + luma_input_stride_y); |
| 1807 | uchar8 cb = vload8(0, in_u.ptr); |
| 1808 | uchar8 cr = vload8(0, in_v.ptr); |
| 1809 | uchar16 cbcr = (uchar16)(cb.s0, cr.s0, cb.s1, cr.s1, cb.s2, cr.s2, cb.s3, cr.s3, cb.s4, cr.s4, cb.s5, cr.s5, cb.s6, |
| 1810 | cr.s6, cb.s7, cr.s7); |
| 1811 | |
| 1812 | vstore16(luma_0, 0, out_y.ptr); |
| 1813 | vstore16(luma_1, 0, out_y.ptr + luma_output_stride_y); |
| 1814 | vstore16(cbcr, 0, out_uv.ptr); |
| 1815 | } |
| 1816 | |
| 1817 | /** Convert a YUYV image to NV12 using BT709 color space |
| 1818 | * |
| 1819 | * Global Workgroup Size [ DIV_CEIL(width, 8), height ] |
| 1820 | * No offset. |
| 1821 | * |
| 1822 | * @param[in] yuyv_input_ptr Pointer to the source image. Supported Format: U8 |
| 1823 | * @param[in] yuyv_input_stride_x Stride of the source image in X dimension (in bytes) |
| 1824 | * @param[in] yuyv_input_step_x yuyv_input_stride_x * number of elements along X processed per workitem(in bytes) |
| 1825 | * @param[in] yuyv_input_stride_y Stride of the source image in Y dimension (in bytes) |
| 1826 | * @param[in] yuyv_input_step_y yuyv_input_stride_y * number of elements along Y processed per workitem(in bytes) |
| 1827 | * @param[in] yuyv_input_offset_first_element_in_bytes The offset of the first element in the source image |
| 1828 | * @param[out] luma_output_ptr Pointer to the destination luma channel. Supported Format: U8 |
| 1829 | * @param[in] luma_output_stride_x Stride of the destination luma channel in X dimension (in bytes) |
| 1830 | * @param[in] luma_output_step_x luma_output_stride_x * number of elements along X processed per workitem(in bytes) |
| 1831 | * @param[in] luma_output_stride_y Stride of the destination image luma channel in Y dimension (in bytes) |
| 1832 | * @param[in] luma_output_step_y luma_output_stride_y * number of elements along Y processed per workitem(in bytes) |
| 1833 | * @param[in] luma_output_offset_first_element_in_bytes The offset of the first element in the destination luma channel |
| 1834 | * @param[out] uv_output_ptr Pointer to the destination UV channel. Supported Format: U8 |
| 1835 | * @param[in] uv_output_stride_x Stride of the destination UV channel in X dimension (in bytes) |
| 1836 | * @param[in] uv_output_step_x uv_output_stride_x * number of elements along X processed per workitem(in bytes) |
| 1837 | * @param[in] uv_output_stride_y Stride of the destination image UV channel in Y dimension (in bytes) |
| 1838 | * @param[in] uv_output_step_y uv_output_stride_y * number of elements along Y processed per workitem(in bytes) |
| 1839 | * @param[in] uv_output_offset_first_element_in_bytes The offset of the first element in the destination UV channel |
| 1840 | * |
| 1841 | */ |
| 1842 | __kernel void YUYV422_to_NV12_bt709( |
| 1843 | IMAGE_DECLARATION(yuyv_input), |
| 1844 | IMAGE_DECLARATION(luma_output), |
| 1845 | IMAGE_DECLARATION(uv_output)) |
| 1846 | { |
| 1847 | Image in_yuyv = CONVERT_TO_IMAGE_STRUCT(yuyv_input); |
| 1848 | Image out_y = CONVERT_TO_IMAGE_STRUCT(luma_output); |
| 1849 | Image out_uv = CONVERT_TO_IMAGE_STRUCT(uv_output); |
| 1850 | |
| 1851 | // handle 16 pixels every time, each line 8 pixels |
| 1852 | uchar16 yuyv = vload16(0, in_yuyv.ptr); |
| 1853 | ushort8 cbcr_0 = (ushort8)(yuyv.s1, yuyv.s3, yuyv.s5, yuyv.s7, yuyv.s9, yuyv.sb, yuyv.sd, yuyv.sf); |
| 1854 | uchar8 luma = (uchar8)(yuyv.s0, yuyv.s2, yuyv.s4, yuyv.s6, yuyv.s8, yuyv.sa, yuyv.sc, yuyv.se); |
| 1855 | vstore8(luma, 0, out_y.ptr); |
| 1856 | |
| 1857 | yuyv = vload16(0, in_yuyv.ptr + yuyv_input_stride_y); |
| 1858 | ushort8 cbcr_1 = (ushort8)(yuyv.s1, yuyv.s3, yuyv.s5, yuyv.s7, yuyv.s9, yuyv.sb, yuyv.sd, yuyv.sf); |
| 1859 | luma = (uchar8)(yuyv.s0, yuyv.s2, yuyv.s4, yuyv.s6, yuyv.s8, yuyv.sa, yuyv.sc, yuyv.se); |
| 1860 | vstore8(luma, 0, out_y.ptr + luma_output_stride_y); |
| 1861 | |
| 1862 | uchar8 cbcr = convert_uchar8((cbcr_0 + cbcr_1) / (ushort8)(2)); |
| 1863 | vstore8(cbcr, 0, out_uv.ptr); |
| 1864 | } |
| 1865 | |
| 1866 | /** Convert a UYVY image to NV12 using BT709 color space |
| 1867 | * |
| 1868 | * Global Workgroup Size [ DIV_CEIL(width, 4), height ] |
| 1869 | * No offset. |
| 1870 | * |
| 1871 | * @param[in] input_uyvy_ptr Pointer to the source image. Supported Format: U8 |
| 1872 | * @param[in] input_uyvy_stride_x Stride of the source image in X dimension (in bytes) |
| 1873 | * @param[in] input_uyvy_step_x input_uyvy_stride_x * number of elements along X processed per workitem(in bytes) |
| 1874 | * @param[in] input_uyvy_stride_y Stride of the source image in Y dimension (in bytes) |
| 1875 | * @param[in] input_uyvy_step_y input_uyvy_stride_y * number of elements along Y processed per workitem(in bytes) |
| 1876 | * @param[in] input_uyvy_offset_first_element_in_bytes The offset of the first element in the source image |
| 1877 | * @param[out] luma_ptr Pointer to the destination luma channel. Supported Format: U8 |
| 1878 | * @param[in] luma_stride_x Stride of the destination luma channel in X dimension (in bytes) |
| 1879 | * @param[in] luma_step_x luma_stride_x * number of elements along X processed per workitem(in bytes) |
| 1880 | * @param[in] luma_stride_y Stride of the destination image luma channel in Y dimension (in bytes) |
| 1881 | * @param[in] luma_step_y luma_stride_y * number of elements along Y processed per workitem(in bytes) |
| 1882 | * @param[in] luma_offset_first_element_in_bytes The offset of the first element in the destination image luma channel |
| 1883 | * @param[out] uv_ptr Pointer to the destination uv channel. Supported Format: U8 |
| 1884 | * @param[in] uv_stride_x Stride of the destination uv channel in X dimension (in bytes) |
| 1885 | * @param[in] uv_step_x uv_stride_x * number of elements along X processed per workitem(in bytes) |
| 1886 | * @param[in] uv_stride_y Stride of the destination image luma channel in Y dimension (in bytes) |
| 1887 | * @param[in] uv_step_y uv_stride_y * number of elements along Y processed per workitem(in bytes) |
| 1888 | * @param[in] uv_offset_first_element_in_bytes The offset of the first element in the destination image uv channel |
| 1889 | * |
| 1890 | */ |
| 1891 | __kernel void UYVY422_to_NV12_bt709( |
| 1892 | IMAGE_DECLARATION(input_uyvy), |
| 1893 | IMAGE_DECLARATION(luma), |
| 1894 | IMAGE_DECLARATION(uv)) |
| 1895 | { |
| 1896 | Image in = CONVERT_TO_IMAGE_STRUCT(input_uyvy); |
| 1897 | Image out_y = CONVERT_TO_IMAGE_STRUCT(luma); |
| 1898 | Image out_uv = CONVERT_TO_IMAGE_STRUCT(uv); |
| 1899 | |
| 1900 | // handle 16 pixels every time, each line 8 pixels |
| 1901 | const uchar16 uyvy_t = vload16(0, in.ptr); |
| 1902 | vstore8(uyvy_t.s13579bdf, 0, out_y.ptr); |
| 1903 | |
| 1904 | const uchar16 uyvy_b = vload16(0, in.ptr + input_uyvy_stride_y); |
| 1905 | vstore8(uyvy_b.s13579bdf, 0, out_y.ptr + luma_stride_y); |
| 1906 | |
| 1907 | const ushort8 cbcr_t = (ushort8)(uyvy_t.s0, uyvy_t.s2, uyvy_t.s4, uyvy_t.s6, uyvy_t.s8, uyvy_t.sa, uyvy_t.sc, uyvy_t.se); |
| 1908 | const ushort8 cbcr_b = (ushort8)(uyvy_b.s0, uyvy_b.s2, uyvy_b.s4, uyvy_b.s6, uyvy_b.s8, uyvy_b.sa, uyvy_b.sc, uyvy_b.se); |
| 1909 | const uchar8 cbcr = convert_uchar8((cbcr_t + cbcr_b) / (ushort8)(2)); |
| 1910 | vstore8(cbcr, 0, out_uv.ptr); |
| 1911 | } |