Line data Source code
1 : /* The copyright in this software is being made available under the BSD 2 : * License, included below. This software may be subject to other third party 3 : * and contributor rights, including patent rights, and no such rights are 4 : * granted under this license. 5 : * 6 : * Copyright 2010-2014, ITU/ISO/IEC 7 : * All rights reserved. 8 : * 9 : * Redistribution and use in source and binary forms, with or without 10 : * modification, are permitted provided that the following conditions are met: 11 : * 12 : * * Redistributions of source code must retain the above copyright notice, 13 : * this list of conditions and the following disclaimer. 14 : * * Redistributions in binary form must reproduce the above copyright notice, 15 : * this list of conditions and the following disclaimer in the documentation 16 : * and/or other materials provided with the distribution. 17 : * * Neither the name of the ITU/ISO/IEC nor the names of its contributors may 18 : * be used to endorse or promote products derived from this software without 19 : * specific prior written permission. 20 : * 21 : * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS" 22 : * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE 23 : * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE 24 : * ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT HOLDER OR CONTRIBUTORS 25 : * BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR 26 : * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF 27 : * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS 28 : * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN 29 : * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) 30 : * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF 31 : * THE POSSIBILITY OF SUCH DAMAGE. 32 : */ 33 : 34 : #include "EbHmCode.h" 35 : #include "EbUtility.h" 36 : 37 : /******************************************* 38 : * compute4x4_satd 39 : * returns 4x4 Sum of Absolute Transformed Differences 40 : *******************************************/ 41 0 : uint64_t compute4x4_satd( 42 : int16_t *diff) // input parameter, diff samples Ptr 43 : { 44 0 : uint64_t satdBlock4x4 = 0; 45 : int16_t m[16], d[16]; 46 : uint32_t k; 47 : 48 : /*===== hadamard transform =====*/ 49 0 : m[0] = diff[0] + diff[12]; 50 0 : m[1] = diff[1] + diff[13]; 51 0 : m[2] = diff[2] + diff[14]; 52 0 : m[3] = diff[3] + diff[15]; 53 0 : m[4] = diff[4] + diff[8]; 54 0 : m[5] = diff[5] + diff[9]; 55 0 : m[6] = diff[6] + diff[10]; 56 0 : m[7] = diff[7] + diff[11]; 57 0 : m[8] = diff[4] - diff[8]; 58 0 : m[9] = diff[5] - diff[9]; 59 0 : m[10] = diff[6] - diff[10]; 60 0 : m[11] = diff[7] - diff[11]; 61 0 : m[12] = diff[0] - diff[12]; 62 0 : m[13] = diff[1] - diff[13]; 63 0 : m[14] = diff[2] - diff[14]; 64 0 : m[15] = diff[3] - diff[15]; 65 : 66 0 : d[0] = m[0] + m[4]; 67 0 : d[1] = m[1] + m[5]; 68 0 : d[2] = m[2] + m[6]; 69 0 : d[3] = m[3] + m[7]; 70 0 : d[4] = m[8] + m[12]; 71 0 : d[5] = m[9] + m[13]; 72 0 : d[6] = m[10] + m[14]; 73 0 : d[7] = m[11] + m[15]; 74 0 : d[8] = m[0] - m[4]; 75 0 : d[9] = m[1] - m[5]; 76 0 : d[10] = m[2] - m[6]; 77 0 : d[11] = m[3] - m[7]; 78 0 : d[12] = m[12] - m[8]; 79 0 : d[13] = m[13] - m[9]; 80 0 : d[14] = m[14] - m[10]; 81 0 : d[15] = m[15] - m[11]; 82 : 83 0 : m[0] = d[0] + d[3]; 84 0 : m[1] = d[1] + d[2]; 85 0 : m[2] = d[1] - d[2]; 86 0 : m[3] = d[0] - d[3]; 87 0 : m[4] = d[4] + d[7]; 88 0 : m[5] = d[5] + d[6]; 89 0 : m[6] = d[5] - d[6]; 90 0 : m[7] = d[4] - d[7]; 91 0 : m[8] = d[8] + d[11]; 92 0 : m[9] = d[9] + d[10]; 93 0 : m[10] = d[9] - d[10]; 94 0 : m[11] = d[8] - d[11]; 95 0 : m[12] = d[12] + d[15]; 96 0 : m[13] = d[13] + d[14]; 97 0 : m[14] = d[13] - d[14]; 98 0 : m[15] = d[12] - d[15]; 99 : 100 0 : d[0] = m[0] + m[1]; 101 0 : d[1] = m[0] - m[1]; 102 0 : d[2] = m[2] + m[3]; 103 0 : d[3] = m[3] - m[2]; 104 0 : d[4] = m[4] + m[5]; 105 0 : d[5] = m[4] - m[5]; 106 0 : d[6] = m[6] + m[7]; 107 0 : d[7] = m[7] - m[6]; 108 0 : d[8] = m[8] + m[9]; 109 0 : d[9] = m[8] - m[9]; 110 0 : d[10] = m[10] + m[11]; 111 0 : d[11] = m[11] - m[10]; 112 0 : d[12] = m[12] + m[13]; 113 0 : d[13] = m[12] - m[13]; 114 0 : d[14] = m[14] + m[15]; 115 0 : d[15] = m[15] - m[14]; 116 : 117 0 : for (k = 0; k < 16; ++k) 118 0 : satdBlock4x4 += ABS(d[k]); 119 0 : satdBlock4x4 = ((satdBlock4x4 + 1) >> 1); 120 : 121 0 : return satdBlock4x4; 122 : } 123 : 124 0 : uint64_t compute4x4_satd_u8( 125 : uint8_t *src, // input parameter, diff samples Ptr 126 : uint64_t *dc_value, 127 : uint32_t src_stride) 128 : { 129 0 : uint64_t satdBlock4x4 = 0; 130 : int16_t m[16], d[16]; 131 : uint32_t k; 132 : 133 : /*===== hadamard transform =====*/ 134 0 : m[0] = src[0] + src[3 * src_stride]; 135 0 : m[1] = src[1] + src[1 + (3 * src_stride)]; 136 0 : m[2] = src[2] + src[2 + (3 * src_stride)]; 137 0 : m[3] = src[3] + src[3 + (3 * src_stride)]; 138 0 : m[4] = src[src_stride] + src[2 * src_stride]; 139 0 : m[5] = src[1 + src_stride] + src[1 + (2 * src_stride)]; 140 0 : m[6] = src[2 + src_stride] + src[2 + (2 * src_stride)]; 141 0 : m[7] = src[3 + src_stride] + src[3 + (2 * src_stride)]; 142 0 : m[8] = src[src_stride] - src[2 * src_stride]; 143 0 : m[9] = src[1 + src_stride] - src[1 + (2 * src_stride)]; 144 0 : m[10] = src[2 + src_stride] - src[2 + (2 * src_stride)]; 145 0 : m[11] = src[3 + src_stride] - src[3 + (2 * src_stride)]; 146 0 : m[12] = src[0] - src[3 * src_stride]; 147 0 : m[13] = src[1] - src[1 + (3 * src_stride)]; 148 0 : m[14] = src[2] - src[2 + (3 * src_stride)]; 149 0 : m[15] = src[3] - src[3 + (3 * src_stride)]; 150 : 151 0 : d[0] = m[0] + m[4]; 152 0 : d[1] = m[1] + m[5]; 153 0 : d[2] = m[2] + m[6]; 154 0 : d[3] = m[3] + m[7]; 155 0 : d[4] = m[8] + m[12]; 156 0 : d[5] = m[9] + m[13]; 157 0 : d[6] = m[10] + m[14]; 158 0 : d[7] = m[11] + m[15]; 159 0 : d[8] = m[0] - m[4]; 160 0 : d[9] = m[1] - m[5]; 161 0 : d[10] = m[2] - m[6]; 162 0 : d[11] = m[3] - m[7]; 163 0 : d[12] = m[12] - m[8]; 164 0 : d[13] = m[13] - m[9]; 165 0 : d[14] = m[14] - m[10]; 166 0 : d[15] = m[15] - m[11]; 167 : 168 0 : m[0] = d[0] + d[3]; 169 0 : m[1] = d[1] + d[2]; 170 0 : m[2] = d[1] - d[2]; 171 0 : m[3] = d[0] - d[3]; 172 0 : m[4] = d[4] + d[7]; 173 0 : m[5] = d[5] + d[6]; 174 0 : m[6] = d[5] - d[6]; 175 0 : m[7] = d[4] - d[7]; 176 0 : m[8] = d[8] + d[11]; 177 0 : m[9] = d[9] + d[10]; 178 0 : m[10] = d[9] - d[10]; 179 0 : m[11] = d[8] - d[11]; 180 0 : m[12] = d[12] + d[15]; 181 0 : m[13] = d[13] + d[14]; 182 0 : m[14] = d[13] - d[14]; 183 0 : m[15] = d[12] - d[15]; 184 : 185 0 : d[0] = m[0] + m[1]; 186 0 : d[1] = m[0] - m[1]; 187 0 : d[2] = m[2] + m[3]; 188 0 : d[3] = m[3] - m[2]; 189 0 : d[4] = m[4] + m[5]; 190 0 : d[5] = m[4] - m[5]; 191 0 : d[6] = m[6] + m[7]; 192 0 : d[7] = m[7] - m[6]; 193 0 : d[8] = m[8] + m[9]; 194 0 : d[9] = m[8] - m[9]; 195 0 : d[10] = m[10] + m[11]; 196 0 : d[11] = m[11] - m[10]; 197 0 : d[12] = m[12] + m[13]; 198 0 : d[13] = m[12] - m[13]; 199 0 : d[14] = m[14] + m[15]; 200 0 : d[15] = m[15] - m[14]; 201 : 202 0 : for (k = 0; k < 16; ++k) 203 0 : satdBlock4x4 += ABS(d[k]); 204 0 : satdBlock4x4 = ((satdBlock4x4 + 1) >> 1); 205 0 : *dc_value += d[0]; 206 0 : return satdBlock4x4; 207 : } 208 : 209 0 : uint64_t compute8x8_satd_u8_c( 210 : uint8_t* src, // input parameter, diff samples Ptr 211 : uint64_t* dcValue, 212 : uint32_t srcStride) 213 : { 214 0 : uint64_t satdBlock8x8 = 0; 215 : int16_t m1[8][8], m2[8][8], m3[8][8]; 216 : uint32_t i, j; 217 : 218 : // Horizontal 219 0 : for (j = 0; j < 8; j++) { 220 0 : m2[j][0] = src[j * srcStride] + src[j * srcStride + 4]; 221 0 : m2[j][1] = src[j * srcStride + 1] + src[j * srcStride + 5]; 222 0 : m2[j][2] = src[j * srcStride + 2] + src[j * srcStride + 6]; 223 0 : m2[j][3] = src[j * srcStride + 3] + src[j * srcStride + 7]; 224 0 : m2[j][4] = src[j * srcStride] - src[j * srcStride + 4]; 225 0 : m2[j][5] = src[j * srcStride + 1] - src[j * srcStride + 5]; 226 0 : m2[j][6] = src[j * srcStride + 2] - src[j * srcStride + 6]; 227 0 : m2[j][7] = src[j * srcStride + 3] - src[j * srcStride + 7]; 228 : 229 0 : m1[j][0] = m2[j][0] + m2[j][2]; 230 0 : m1[j][1] = m2[j][1] + m2[j][3]; 231 0 : m1[j][2] = m2[j][0] - m2[j][2]; 232 0 : m1[j][3] = m2[j][1] - m2[j][3]; 233 0 : m1[j][4] = m2[j][4] + m2[j][6]; 234 0 : m1[j][5] = m2[j][5] + m2[j][7]; 235 0 : m1[j][6] = m2[j][4] - m2[j][6]; 236 0 : m1[j][7] = m2[j][5] - m2[j][7]; 237 : 238 0 : m2[j][0] = m1[j][0] + m1[j][1]; 239 0 : m2[j][1] = m1[j][0] - m1[j][1]; 240 0 : m2[j][2] = m1[j][2] + m1[j][3]; 241 0 : m2[j][3] = m1[j][2] - m1[j][3]; 242 0 : m2[j][4] = m1[j][4] + m1[j][5]; 243 0 : m2[j][5] = m1[j][4] - m1[j][5]; 244 0 : m2[j][6] = m1[j][6] + m1[j][7]; 245 0 : m2[j][7] = m1[j][6] - m1[j][7]; 246 : } 247 : 248 : // Vertical 249 0 : for (i = 0; i < 8; i++) { 250 0 : m3[0][i] = m2[0][i] + m2[4][i]; 251 0 : m3[1][i] = m2[1][i] + m2[5][i]; 252 0 : m3[2][i] = m2[2][i] + m2[6][i]; 253 0 : m3[3][i] = m2[3][i] + m2[7][i]; 254 0 : m3[4][i] = m2[0][i] - m2[4][i]; 255 0 : m3[5][i] = m2[1][i] - m2[5][i]; 256 0 : m3[6][i] = m2[2][i] - m2[6][i]; 257 0 : m3[7][i] = m2[3][i] - m2[7][i]; 258 : 259 0 : m1[0][i] = m3[0][i] + m3[2][i]; 260 0 : m1[1][i] = m3[1][i] + m3[3][i]; 261 0 : m1[2][i] = m3[0][i] - m3[2][i]; 262 0 : m1[3][i] = m3[1][i] - m3[3][i]; 263 0 : m1[4][i] = m3[4][i] + m3[6][i]; 264 0 : m1[5][i] = m3[5][i] + m3[7][i]; 265 0 : m1[6][i] = m3[4][i] - m3[6][i]; 266 0 : m1[7][i] = m3[5][i] - m3[7][i]; 267 : 268 0 : m2[0][i] = m1[0][i] + m1[1][i]; 269 0 : m2[1][i] = m1[0][i] - m1[1][i]; 270 0 : m2[2][i] = m1[2][i] + m1[3][i]; 271 0 : m2[3][i] = m1[2][i] - m1[3][i]; 272 0 : m2[4][i] = m1[4][i] + m1[5][i]; 273 0 : m2[5][i] = m1[4][i] - m1[5][i]; 274 0 : m2[6][i] = m1[6][i] + m1[7][i]; 275 0 : m2[7][i] = m1[6][i] - m1[7][i]; 276 : } 277 : 278 0 : for (i = 0; i < 8; i++) { 279 0 : for (j = 0; j < 8; j++) { 280 0 : satdBlock8x8 += (uint64_t)ABS(m2[i][j]); 281 : } 282 : } 283 : 284 0 : satdBlock8x8 = ((satdBlock8x8 + 2) >> 2); 285 0 : *dcValue += m2[0][0]; 286 0 : return satdBlock8x8; 287 : }
