--- trunk/xvidcore/src/motion/sad.c 2002/09/06 16:59:47 430 +++ branches/dev-api-4/xvidcore/src/motion/sad.c 2003/06/09 01:25:19 1053 @@ -1,91 +1,58 @@ -/***************************************************************************** +/************************************************************************** * - * XVID MPEG-4 VIDEO CODEC - * - SAD calculation module (C part) - + * XVID MPEG-4 VIDEO CODEC + * sum of absolute difference * - * Copyright(C) 2002 Michael Militzer - * Copyright(C) 2002 Pete Ross + * This program is an implementation of a part of one or more MPEG-4 + * Video tools as specified in ISO/IEC 14496-2 standard. Those intending + * to use this software module in hardware or software products are + * advised that its use may infringe existing patents or copyrights, and + * any such use would be at such party's own risk. The original + * developer of this software module and his/her company, and subsequent + * editors and their companies, will have no liability for use of this + * software or modifications or derivatives thereof. * - * This program is an implementation of a part of one or more MPEG-4 - * Video tools as specified in ISO/IEC 14496-2 standard. Those intending - * to use this software module in hardware or software products are - * advised that its use may infringe existing patents or copyrights, and - * any such use would be at such party's own risk. The original - * developer of this software module and his/her company, and subsequent - * editors and their companies, will have no liability for use of this - * software or modifications or derivatives thereof. + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. * - * This program is free software; you can redistribute it and/or modify - * it under the terms of the GNU General Public License as published by - * the Free Software Foundation; either version 2 of the License, or - * (at your option) any later version. + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. * - * This program is distributed in the hope that it will be useful, - * but WITHOUT ANY WARRANTY; without even the implied warranty of - * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the - * GNU General Public License for more details. + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA. * - * You should have received a copy of the GNU General Public License - * along with this program; if not, write to the Free Software - * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA + *************************************************************************/ + +/************************************************************************** + * + * History: * - ****************************************************************************/ + * 14.02.2002 added sad16bi_c() + * 10.11.2001 initial version; (c)2001 peter ross + * + *************************************************************************/ + #include "../portab.h" +#include "../global.h" #include "sad.h" +#include + sad16FuncPtr sad16; sad8FuncPtr sad8; sad16biFuncPtr sad16bi; -sad8biFuncPtr sad8bi; // not really sad16, but no difference in prototype +sad8biFuncPtr sad8bi; /* not really sad16, but no difference in prototype */ dev16FuncPtr dev16; +sad16vFuncPtr sad16v; sadInitFuncPtr sadInit; -#define ABS(X) (((X)>0)?(X):-(X)) - -#define MRSAD16_CORRFACTOR 8 -uint32_t -mrsad16_c(const uint8_t * const cur, - const uint8_t * const ref, - const uint32_t stride, - const uint32_t best_sad) -{ - - uint32_t sad = 0; - int32_t mean = 0; - uint32_t i, j; - uint8_t const *ptr_cur = cur; - uint8_t const *ptr_ref = ref; - - for (j = 0; j < 16; j++) { - for (i = 0; i < 16; i++) { - mean += ((int) *(ptr_cur + i) - (int) *(ptr_ref + i)); - } - ptr_cur += stride; - ptr_ref += stride; - - } - mean /= 256; - - for (j = 0; j < 16; j++) { - - ptr_cur -= stride; - ptr_ref -= stride; - - for (i = 0; i < 16; i++) { - - sad += ABS(*(ptr_cur + i) - *(ptr_ref + i) - mean); - if (sad >= best_sad) { - return MRSAD16_CORRFACTOR * sad; - } - } - } - - return MRSAD16_CORRFACTOR * sad; - -} - uint32_t sad16_c(const uint8_t * const cur, @@ -95,25 +62,33 @@ { uint32_t sad = 0; - uint32_t i, j; + uint32_t j; uint8_t const *ptr_cur = cur; uint8_t const *ptr_ref = ref; for (j = 0; j < 16; j++) { + sad += abs(ptr_cur[0] - ptr_ref[0]); + sad += abs(ptr_cur[1] - ptr_ref[1]); + sad += abs(ptr_cur[2] - ptr_ref[2]); + sad += abs(ptr_cur[3] - ptr_ref[3]); + sad += abs(ptr_cur[4] - ptr_ref[4]); + sad += abs(ptr_cur[5] - ptr_ref[5]); + sad += abs(ptr_cur[6] - ptr_ref[6]); + sad += abs(ptr_cur[7] - ptr_ref[7]); + sad += abs(ptr_cur[8] - ptr_ref[8]); + sad += abs(ptr_cur[9] - ptr_ref[9]); + sad += abs(ptr_cur[10] - ptr_ref[10]); + sad += abs(ptr_cur[11] - ptr_ref[11]); + sad += abs(ptr_cur[12] - ptr_ref[12]); + sad += abs(ptr_cur[13] - ptr_ref[13]); + sad += abs(ptr_cur[14] - ptr_ref[14]); + sad += abs(ptr_cur[15] - ptr_ref[15]); - for (i = 0; i < 16; i++) { - - sad += ABS(*(ptr_cur + i) - *(ptr_ref + i)); - - if (sad >= best_sad) { + if (sad >= best_sad) return sad; - } - - } - - ptr_cur += stride; - ptr_ref += stride; + ptr_cur += stride; + ptr_ref += stride; } @@ -121,8 +96,6 @@ } - - uint32_t sad16bi_c(const uint8_t * const cur, const uint8_t * const ref1, @@ -140,14 +113,7 @@ for (i = 0; i < 16; i++) { int pixel = (ptr_ref1[i] + ptr_ref2[i] + 1) / 2; - - if (pixel < 0) { - pixel = 0; - } else if (pixel > 255) { - pixel = 255; - } - - sad += ABS(ptr_cur[i] - pixel); + sad += abs(ptr_cur[i] - pixel); } ptr_cur += stride; @@ -177,14 +143,7 @@ for (i = 0; i < 8; i++) { int pixel = (ptr_ref1[i] + ptr_ref2[i] + 1) / 2; - - if (pixel < 0) { - pixel = 0; - } else if (pixel > 255) { - pixel = 255; - } - - sad += ABS(ptr_cur[i] - pixel); + sad += abs(ptr_cur[i] - pixel); } ptr_cur += stride; @@ -205,16 +164,21 @@ const uint32_t stride) { uint32_t sad = 0; - uint32_t i, j; + uint32_t j; uint8_t const *ptr_cur = cur; uint8_t const *ptr_ref = ref; for (j = 0; j < 8; j++) { - for (i = 0; i < 8; i++) { - sad += ABS(*(ptr_cur + i) - *(ptr_ref + i)); - } - + sad += abs(ptr_cur[0] - ptr_ref[0]); + sad += abs(ptr_cur[1] - ptr_ref[1]); + sad += abs(ptr_cur[2] - ptr_ref[2]); + sad += abs(ptr_cur[3] - ptr_ref[3]); + sad += abs(ptr_cur[4] - ptr_ref[4]); + sad += abs(ptr_cur[5] - ptr_ref[5]); + sad += abs(ptr_cur[6] - ptr_ref[6]); + sad += abs(ptr_cur[7] - ptr_ref[7]); + ptr_cur += stride; ptr_ref += stride; @@ -224,8 +188,6 @@ } - - /* average deviation from mean */ uint32_t @@ -253,7 +215,7 @@ for (j = 0; j < 16; j++) { for (i = 0; i < 16; i++) - dev += ABS(*(ptr_cur + i) - (int32_t) mean); + dev += abs(*(ptr_cur + i) - (int32_t) mean); ptr_cur += stride; @@ -261,3 +223,75 @@ return dev; } + +uint32_t sad16v_c(const uint8_t * const cur, + const uint8_t * const ref, + const uint32_t stride, + int32_t *sad) +{ + sad[0] = sad8(cur, ref, stride); + sad[1] = sad8(cur + 8, ref + 8, stride); + sad[2] = sad8(cur + 8*stride, ref + 8*stride, stride); + sad[3] = sad8(cur + 8*stride + 8, ref + 8*stride + 8, stride); + + return sad[0]+sad[1]+sad[2]+sad[3]; +} + +uint32_t sad32v_c(const uint8_t * const cur, + const uint8_t * const ref, + const uint32_t stride, + int32_t *sad) +{ + sad[0] = sad16(cur, ref, stride, 256*4096); + sad[1] = sad16(cur + 8, ref + 8, stride, 256*4096); + sad[2] = sad16(cur + 8*stride, ref + 8*stride, stride, 256*4096); + sad[3] = sad16(cur + 8*stride + 8, ref + 8*stride + 8, stride, 256*4096); + + return sad[0]+sad[1]+sad[2]+sad[3]; +} + + + +#define MRSAD16_CORRFACTOR 8 +uint32_t +mrsad16_c(const uint8_t * const cur, + const uint8_t * const ref, + const uint32_t stride, + const uint32_t best_sad) +{ + + uint32_t sad = 0; + int32_t mean = 0; + uint32_t i, j; + uint8_t const *ptr_cur = cur; + uint8_t const *ptr_ref = ref; + + for (j = 0; j < 16; j++) { + for (i = 0; i < 16; i++) { + mean += ((int) *(ptr_cur + i) - (int) *(ptr_ref + i)); + } + ptr_cur += stride; + ptr_ref += stride; + + } + mean /= 256; + + for (j = 0; j < 16; j++) { + + ptr_cur -= stride; + ptr_ref -= stride; + + for (i = 0; i < 16; i++) { + + sad += abs(*(ptr_cur + i) - *(ptr_ref + i) - mean); + if (sad >= best_sad) { + return MRSAD16_CORRFACTOR * sad; + } + } + } + + return MRSAD16_CORRFACTOR * sad; + +} + +