2 |
* |
* |
3 |
* Modifications: |
* Modifications: |
4 |
* |
* |
5 |
|
* 01.05.2002 updated MotionEstimationBVOP |
6 |
* 25.04.2002 partial prevMB conversion |
* 25.04.2002 partial prevMB conversion |
7 |
* 22.04.2002 remove some compile warning by chenm001 <chenm001@163.com> |
* 22.04.2002 remove some compile warning by chenm001 <chenm001@163.com> |
8 |
* 14.04.2002 added MotionEstimationBVOP() |
* 14.04.2002 added MotionEstimationBVOP() |
57 |
/* sad16(0,0) bias; mpeg4 spec suggests nb/2+1 */ |
/* sad16(0,0) bias; mpeg4 spec suggests nb/2+1 */ |
58 |
/* nb = vop pixels * 2^(bpp-8) */ |
/* nb = vop pixels * 2^(bpp-8) */ |
59 |
#define MV16_00_BIAS (128+1) |
#define MV16_00_BIAS (128+1) |
60 |
|
#define MV8_00_BIAS (0) |
61 |
|
|
62 |
/* INTER bias for INTER/INTRA decision; mpeg4 spec suggests 2*nb */ |
/* INTER bias for INTER/INTRA decision; mpeg4 spec suggests 2*nb */ |
63 |
#define INTER_BIAS 512 |
#define INTER_BIAS 512 |
185 |
|
|
186 |
typedef MainSearch8Func* MainSearch8FuncPtr; |
typedef MainSearch8Func* MainSearch8FuncPtr; |
187 |
|
|
188 |
|
static int32_t lambda_vec16[32] = /* rounded values for lambda param for weight of motion bits as in modified H.26L */ |
189 |
|
{ 0 ,(int)(1.00235+0.5), (int)(1.15582+0.5), (int)(1.31976+0.5), (int)(1.49591+0.5), (int)(1.68601+0.5), |
190 |
|
(int)(1.89187+0.5), (int)(2.11542+0.5), (int)(2.35878+0.5), (int)(2.62429+0.5), (int)(2.91455+0.5), |
191 |
|
(int)(3.23253+0.5), (int)(3.58158+0.5), (int)(3.96555+0.5), (int)(4.38887+0.5), (int)(4.85673+0.5), |
192 |
|
(int)(5.37519+0.5), (int)(5.95144+0.5), (int)(6.59408+0.5), (int)(7.31349+0.5), (int)(8.12242+0.5), |
193 |
|
(int)(9.03669+0.5), (int)(10.0763+0.5), (int)(11.2669+0.5), (int)(12.6426+0.5), (int)(14.2493+0.5), |
194 |
|
(int)(16.1512+0.5), (int)(18.442+0.5), (int)(21.2656+0.5), (int)(24.8580+0.5), (int)(29.6436+0.5), |
195 |
|
(int)(36.4949+0.5) }; |
196 |
|
|
197 |
|
static int32_t *lambda_vec8 = lambda_vec16; /* same table for INTER and INTER4V for now*/ |
198 |
|
|
199 |
|
|
200 |
|
|
201 |
// mv.length table |
// mv.length table |
202 |
static const uint32_t mvtab[33] = { |
static const uint32_t mvtab[33] = { |
203 |
1, 2, 3, 4, 6, 7, 7, 7, |
1, 2, 3, 4, 6, 7, 7, 7, |
233 |
} |
} |
234 |
|
|
235 |
|
|
236 |
static __inline uint32_t calc_delta_16(const int32_t dx, const int32_t dy, const uint32_t iFcode) |
static __inline uint32_t calc_delta_16(const int32_t dx, const int32_t dy, const uint32_t iFcode, const uint32_t iQuant) |
237 |
{ |
{ |
238 |
return NEIGH_TEND_16X16 * (mv_bits(dx, iFcode) + mv_bits(dy, iFcode)); |
return NEIGH_TEND_16X16 * lambda_vec16[iQuant] * (mv_bits(dx, iFcode) + mv_bits(dy, iFcode)); |
239 |
} |
} |
240 |
|
|
241 |
static __inline uint32_t calc_delta_8(const int32_t dx, const int32_t dy, const uint32_t iFcode) |
static __inline uint32_t calc_delta_8(const int32_t dx, const int32_t dy, const uint32_t iFcode, const uint32_t iQuant) |
242 |
|
|
243 |
{ |
{ |
244 |
return NEIGH_TEND_8X8 * (mv_bits(dx, iFcode) + mv_bits(dy, iFcode)); |
return NEIGH_TEND_8X8 * lambda_vec8[iQuant] * (mv_bits(dx, iFcode) + mv_bits(dy, iFcode)); |
245 |
} |
} |
246 |
|
|
247 |
|
|
294 |
/* eventhough we have a seperate prevMBs, |
/* eventhough we have a seperate prevMBs, |
295 |
pmvfast/epsz does something "funny" with the previous frames data */ |
pmvfast/epsz does something "funny" with the previous frames data */ |
296 |
|
|
297 |
for (i = 0; i < iHcount; i++) |
/* for (i = 0; i < iHcount; i++) |
298 |
for (j = 0; j < iWcount; j++) |
for (j = 0; j < iWcount; j++) |
299 |
{ |
{ |
300 |
pMBs[j + i * iWcount].mvs[0] = prevMBs[j + i * iWcount].mvs[0]; |
pMBs[j + i * iWcount].mvs[0] = prevMBs[j + i * iWcount].mvs[0]; |
302 |
pMBs[j + i * iWcount].mvs[2] = prevMBs[j + i * iWcount].mvs[2]; |
pMBs[j + i * iWcount].mvs[2] = prevMBs[j + i * iWcount].mvs[2]; |
303 |
pMBs[j + i * iWcount].mvs[3] = prevMBs[j + i * iWcount].mvs[3]; |
pMBs[j + i * iWcount].mvs[3] = prevMBs[j + i * iWcount].mvs[3]; |
304 |
} |
} |
305 |
|
*/ |
306 |
/*dprintf("*** BEFORE ***"); |
/*dprintf("*** BEFORE ***"); |
307 |
for (i = 0; i < iHcount; i++) |
for (i = 0; i < iHcount; i++) |
308 |
for (j = 0; j < iWcount; j++) |
for (j = 0; j < iWcount; j++) |
445 |
&& (0 <= max_dy) && (0 >= min_dy) ) \ |
&& (0 <= max_dy) && (0 >= min_dy) ) \ |
446 |
{ \ |
{ \ |
447 |
iSAD = sad16( cur, get_ref(pRef, pRefH, pRefV, pRefHV, x, y, 16, 0, 0 , iEdgedWidth), iEdgedWidth, MV_MAX_ERROR); \ |
iSAD = sad16( cur, get_ref(pRef, pRefH, pRefV, pRefHV, x, y, 16, 0, 0 , iEdgedWidth), iEdgedWidth, MV_MAX_ERROR); \ |
448 |
iSAD += calc_delta_16(-pmv[0].x, -pmv[0].y, (uint8_t)iFcode) * iQuant;\ |
iSAD += calc_delta_16(-pmv[0].x, -pmv[0].y, (uint8_t)iFcode, iQuant);\ |
|
if (iSAD <= iQuant * 96) \ |
|
|
iSAD -= MV16_00_BIAS; \ |
|
449 |
if (iSAD < iMinSAD) \ |
if (iSAD < iMinSAD) \ |
450 |
{ iMinSAD=iSAD; currMV->x=0; currMV->y=0; } } \ |
{ iMinSAD=iSAD; currMV->x=0; currMV->y=0; } } \ |
451 |
} |
} |
452 |
|
|
453 |
#define NOCHECK_MV16_CANDIDATE(X,Y) { \ |
#define NOCHECK_MV16_CANDIDATE(X,Y) { \ |
454 |
iSAD = sad16( cur, get_ref(pRef, pRefH, pRefV, pRefHV, x, y, 16, X, Y, iEdgedWidth),iEdgedWidth, iMinSAD); \ |
iSAD = sad16( cur, get_ref(pRef, pRefH, pRefV, pRefHV, x, y, 16, X, Y, iEdgedWidth),iEdgedWidth, iMinSAD); \ |
455 |
iSAD += calc_delta_16((X) - pmv[0].x, (Y) - pmv[0].y, (uint8_t)iFcode) * iQuant;\ |
iSAD += calc_delta_16((X) - pmv[0].x, (Y) - pmv[0].y, (uint8_t)iFcode, iQuant);\ |
456 |
if (iSAD < iMinSAD) \ |
if (iSAD < iMinSAD) \ |
457 |
{ iMinSAD=iSAD; currMV->x=(X); currMV->y=(Y); } \ |
{ iMinSAD=iSAD; currMV->x=(X); currMV->y=(Y); } \ |
458 |
} |
} |
462 |
&& ((Y) <= max_dy) && ((Y) >= min_dy) ) \ |
&& ((Y) <= max_dy) && ((Y) >= min_dy) ) \ |
463 |
{ \ |
{ \ |
464 |
iSAD = sad16( cur, get_ref(pRef, pRefH, pRefV, pRefHV, x, y, 16, X, Y, iEdgedWidth),iEdgedWidth, iMinSAD); \ |
iSAD = sad16( cur, get_ref(pRef, pRefH, pRefV, pRefHV, x, y, 16, X, Y, iEdgedWidth),iEdgedWidth, iMinSAD); \ |
465 |
iSAD += calc_delta_16((X) - pmv[0].x, (Y) - pmv[0].y, (uint8_t)iFcode) * iQuant;\ |
iSAD += calc_delta_16((X) - pmv[0].x, (Y) - pmv[0].y, (uint8_t)iFcode, iQuant);\ |
466 |
if (iSAD < iMinSAD) \ |
if (iSAD < iMinSAD) \ |
467 |
{ iMinSAD=iSAD; currMV->x=(X); currMV->y=(Y); } } \ |
{ iMinSAD=iSAD; currMV->x=(X); currMV->y=(Y); } } \ |
468 |
} |
} |
472 |
&& ((Y) <= max_dy) && ((Y) >= min_dy) ) \ |
&& ((Y) <= max_dy) && ((Y) >= min_dy) ) \ |
473 |
{ \ |
{ \ |
474 |
iSAD = sad16( cur, get_ref(pRef, pRefH, pRefV, pRefHV, x, y, 16, X, Y, iEdgedWidth),iEdgedWidth, iMinSAD); \ |
iSAD = sad16( cur, get_ref(pRef, pRefH, pRefV, pRefHV, x, y, 16, X, Y, iEdgedWidth),iEdgedWidth, iMinSAD); \ |
475 |
iSAD += calc_delta_16((X) - pmv[0].x, (Y) - pmv[0].y, (uint8_t)iFcode) * iQuant;\ |
iSAD += calc_delta_16((X) - pmv[0].x, (Y) - pmv[0].y, (uint8_t)iFcode, iQuant);\ |
476 |
if (iSAD < iMinSAD) \ |
if (iSAD < iMinSAD) \ |
477 |
{ iMinSAD=iSAD; currMV->x=(X); currMV->y=(Y); iDirection=(D); } } \ |
{ iMinSAD=iSAD; currMV->x=(X); currMV->y=(Y); iDirection=(D); } } \ |
478 |
} |
} |
482 |
&& ((Y) <= max_dy) && ((Y) >= min_dy) ) \ |
&& ((Y) <= max_dy) && ((Y) >= min_dy) ) \ |
483 |
{ \ |
{ \ |
484 |
iSAD = sad16( cur, get_ref(pRef, pRefH, pRefV, pRefHV, x, y, 16, X, Y, iEdgedWidth),iEdgedWidth, iMinSAD); \ |
iSAD = sad16( cur, get_ref(pRef, pRefH, pRefV, pRefHV, x, y, 16, X, Y, iEdgedWidth),iEdgedWidth, iMinSAD); \ |
485 |
iSAD += calc_delta_16((X) - pmv[0].x, (Y) - pmv[0].y, (uint8_t)iFcode) * iQuant;\ |
iSAD += calc_delta_16((X) - pmv[0].x, (Y) - pmv[0].y, (uint8_t)iFcode, iQuant);\ |
486 |
if (iSAD < iMinSAD) \ |
if (iSAD < iMinSAD) \ |
487 |
{ iMinSAD=iSAD; currMV->x=(X); currMV->y=(Y); iDirection=(D); iFound=0; } } \ |
{ iMinSAD=iSAD; currMV->x=(X); currMV->y=(Y); iDirection=(D); iFound=0; } } \ |
488 |
} |
} |
490 |
|
|
491 |
#define CHECK_MV8_ZERO {\ |
#define CHECK_MV8_ZERO {\ |
492 |
iSAD = sad8( cur, get_ref(pRef, pRefH, pRefV, pRefHV, x, y, 8, 0, 0 , iEdgedWidth), iEdgedWidth); \ |
iSAD = sad8( cur, get_ref(pRef, pRefH, pRefV, pRefHV, x, y, 8, 0, 0 , iEdgedWidth), iEdgedWidth); \ |
493 |
iSAD += calc_delta_8(-pmv[0].x, -pmv[0].y, (uint8_t)iFcode) * iQuant;\ |
iSAD += calc_delta_8(-pmv[0].x, -pmv[0].y, (uint8_t)iFcode, iQuant);\ |
494 |
if (iSAD < iMinSAD) \ |
if (iSAD < iMinSAD) \ |
495 |
{ iMinSAD=iSAD; currMV->x=0; currMV->y=0; } \ |
{ iMinSAD=iSAD; currMV->x=0; currMV->y=0; } \ |
496 |
} |
} |
498 |
#define NOCHECK_MV8_CANDIDATE(X,Y) \ |
#define NOCHECK_MV8_CANDIDATE(X,Y) \ |
499 |
{ \ |
{ \ |
500 |
iSAD = sad8( cur, get_ref(pRef, pRefH, pRefV, pRefHV, x, y, 8, (X), (Y), iEdgedWidth),iEdgedWidth); \ |
iSAD = sad8( cur, get_ref(pRef, pRefH, pRefV, pRefHV, x, y, 8, (X), (Y), iEdgedWidth),iEdgedWidth); \ |
501 |
iSAD += calc_delta_8((X)-pmv[0].x, (Y)-pmv[0].y, (uint8_t)iFcode) * iQuant;\ |
iSAD += calc_delta_8((X)-pmv[0].x, (Y)-pmv[0].y, (uint8_t)iFcode, iQuant);\ |
502 |
if (iSAD < iMinSAD) \ |
if (iSAD < iMinSAD) \ |
503 |
{ iMinSAD=iSAD; currMV->x=(X); currMV->y=(Y); } \ |
{ iMinSAD=iSAD; currMV->x=(X); currMV->y=(Y); } \ |
504 |
} |
} |
508 |
&& ((Y) <= max_dy) && ((Y) >= min_dy) ) \ |
&& ((Y) <= max_dy) && ((Y) >= min_dy) ) \ |
509 |
{ \ |
{ \ |
510 |
iSAD = sad8( cur, get_ref(pRef, pRefH, pRefV, pRefHV, x, y, 8, (X), (Y), iEdgedWidth),iEdgedWidth); \ |
iSAD = sad8( cur, get_ref(pRef, pRefH, pRefV, pRefHV, x, y, 8, (X), (Y), iEdgedWidth),iEdgedWidth); \ |
511 |
iSAD += calc_delta_8((X)-pmv[0].x, (Y)-pmv[0].y, (uint8_t)iFcode) * iQuant;\ |
iSAD += calc_delta_8((X)-pmv[0].x, (Y)-pmv[0].y, (uint8_t)iFcode, iQuant);\ |
512 |
if (iSAD < iMinSAD) \ |
if (iSAD < iMinSAD) \ |
513 |
{ iMinSAD=iSAD; currMV->x=(X); currMV->y=(Y); } } \ |
{ iMinSAD=iSAD; currMV->x=(X); currMV->y=(Y); } } \ |
514 |
} |
} |
518 |
&& ((Y) <= max_dy) && ((Y) >= min_dy) ) \ |
&& ((Y) <= max_dy) && ((Y) >= min_dy) ) \ |
519 |
{ \ |
{ \ |
520 |
iSAD = sad8( cur, get_ref(pRef, pRefH, pRefV, pRefHV, x, y, 8, (X), (Y), iEdgedWidth),iEdgedWidth); \ |
iSAD = sad8( cur, get_ref(pRef, pRefH, pRefV, pRefHV, x, y, 8, (X), (Y), iEdgedWidth),iEdgedWidth); \ |
521 |
iSAD += calc_delta_8((X)-pmv[0].x, (Y)-pmv[0].y, (uint8_t)iFcode) * iQuant;\ |
iSAD += calc_delta_8((X)-pmv[0].x, (Y)-pmv[0].y, (uint8_t)iFcode, iQuant);\ |
522 |
if (iSAD < iMinSAD) \ |
if (iSAD < iMinSAD) \ |
523 |
{ iMinSAD=iSAD; currMV->x=(X); currMV->y=(Y); iDirection=(D); } } \ |
{ iMinSAD=iSAD; currMV->x=(X); currMV->y=(Y); iDirection=(D); } } \ |
524 |
} |
} |
528 |
&& ((Y) <= max_dy) && ((Y) >= min_dy) ) \ |
&& ((Y) <= max_dy) && ((Y) >= min_dy) ) \ |
529 |
{ \ |
{ \ |
530 |
iSAD = sad8( cur, get_ref(pRef, pRefH, pRefV, pRefHV, x, y, 8, (X), (Y), iEdgedWidth),iEdgedWidth); \ |
iSAD = sad8( cur, get_ref(pRef, pRefH, pRefV, pRefHV, x, y, 8, (X), (Y), iEdgedWidth),iEdgedWidth); \ |
531 |
iSAD += calc_delta_8((X)-pmv[0].x, (Y)-pmv[0].y, (uint8_t)iFcode) * iQuant;\ |
iSAD += calc_delta_8((X)-pmv[0].x, (Y)-pmv[0].y, (uint8_t)iFcode, iQuant);\ |
532 |
if (iSAD < iMinSAD) \ |
if (iSAD < iMinSAD) \ |
533 |
{ iMinSAD=iSAD; currMV->x=(X); currMV->y=(Y); iDirection=(D); iFound=0; } } \ |
{ iMinSAD=iSAD; currMV->x=(X); currMV->y=(Y); iDirection=(D); iFound=0; } } \ |
534 |
} |
} |
1000 |
iMinSAD = sad16( cur, |
iMinSAD = sad16( cur, |
1001 |
get_ref_mv(pRef, pRefH, pRefV, pRefHV, x, y, 16, currMV, iEdgedWidth), |
get_ref_mv(pRef, pRefH, pRefV, pRefHV, x, y, 16, currMV, iEdgedWidth), |
1002 |
iEdgedWidth, MV_MAX_ERROR); |
iEdgedWidth, MV_MAX_ERROR); |
1003 |
iMinSAD += calc_delta_16(currMV->x-pmv[0].x, currMV->y-pmv[0].y, (uint8_t)iFcode) * iQuant; |
iMinSAD += calc_delta_16(currMV->x-pmv[0].x, currMV->y-pmv[0].y, (uint8_t)iFcode, iQuant); |
1004 |
|
|
1005 |
if ( (iMinSAD < 256 ) || ( (MVequal(*currMV,prevMB->mvs[0])) && ((uint32_t)iMinSAD < prevMB->sad16) ) ) |
if ( (iMinSAD < 256 ) || ( (MVequal(*currMV,prevMB->mvs[0])) && ((uint32_t)iMinSAD < prevMB->sad16) ) ) |
1006 |
{ |
{ |
1015 |
Step 5: Calculate SAD for motion vectors taken from left block, top, top-right, and Previous frame block. |
Step 5: Calculate SAD for motion vectors taken from left block, top, top-right, and Previous frame block. |
1016 |
Also calculate (0,0) but do not subtract offset. |
Also calculate (0,0) but do not subtract offset. |
1017 |
Let MinSAD be the smallest SAD up to this point. |
Let MinSAD be the smallest SAD up to this point. |
1018 |
If MV is (0,0) subtract offset. ******** WHAT'S THIS 'OFFSET' ??? *********** |
If MV is (0,0) subtract offset. |
1019 |
*/ |
*/ |
1020 |
|
|
1021 |
// (0,0) is always possible |
// (0,0) is always possible |
1055 |
} |
} |
1056 |
} |
} |
1057 |
|
|
1058 |
|
if ( (MVzero(*currMV)) && (!MVzero(pmv[0])) /* && (iMinSAD <= iQuant * 96)*/ ) |
1059 |
|
iMinSAD -= MV16_00_BIAS; |
1060 |
|
|
1061 |
|
|
1062 |
/* Step 6: If MinSAD <= thresa goto Step 10. |
/* Step 6: If MinSAD <= thresa goto Step 10. |
1063 |
If Motion Vector equal to Previous frame motion vector and MinSAD<PrevFrmSAD goto Step 10. |
If Motion Vector equal to Previous frame motion vector and MinSAD<PrevFrmSAD goto Step 10. |
1064 |
*/ |
*/ |
1291 |
get_range(&min_dx, &max_dx, &min_dy, &max_dy, |
get_range(&min_dx, &max_dx, &min_dy, &max_dy, |
1292 |
x, y, 8, iWidth, iHeight, iFcode); |
x, y, 8, iWidth, iHeight, iFcode); |
1293 |
|
|
|
/* we work with abs. MVs, not relative to prediction, so range is relative to 0,0 */ |
|
|
|
|
1294 |
if (!(MotionFlags & PMV_HALFPELDIAMOND8 )) |
if (!(MotionFlags & PMV_HALFPELDIAMOND8 )) |
1295 |
{ min_dx = EVEN(min_dx); |
{ min_dx = EVEN(min_dx); |
1296 |
max_dx = EVEN(max_dx); |
max_dx = EVEN(max_dx); |
1323 |
If PredEq=1 and MVpredicted = Previous Frame MV, set Found=2 |
If PredEq=1 and MVpredicted = Previous Frame MV, set Found=2 |
1324 |
*/ |
*/ |
1325 |
|
|
1326 |
if ((bPredEq) && (MVequal(pmv[0],pMB->mvs[iSubBlock]) ) ) |
if ((bPredEq) && (MVequal(pmv[0],prevMB->mvs[iSubBlock]) ) ) |
1327 |
iFound=2; |
iFound=2; |
1328 |
|
|
1329 |
/* Step 3: If Distance>0 or thresb<1536 or PredEq=1 Select small Diamond Search. |
/* Step 3: If Distance>0 or thresb<1536 or PredEq=1 Select small Diamond Search. |
1354 |
iMinSAD = sad8( cur, |
iMinSAD = sad8( cur, |
1355 |
get_ref_mv(pRef, pRefH, pRefV, pRefHV, x, y, 8, currMV, iEdgedWidth), |
get_ref_mv(pRef, pRefH, pRefV, pRefHV, x, y, 8, currMV, iEdgedWidth), |
1356 |
iEdgedWidth); |
iEdgedWidth); |
1357 |
iMinSAD += calc_delta_8(currMV->x - pmv[0].x, currMV->y - pmv[0].y, (uint8_t)iFcode) * iQuant; |
iMinSAD += calc_delta_8(currMV->x - pmv[0].x, currMV->y - pmv[0].y, (uint8_t)iFcode, iQuant); |
1358 |
|
|
1359 |
if ( (iMinSAD < 256/4 ) || ( (MVequal(*currMV,pMB->mvs[iSubBlock])) && ((uint32_t)iMinSAD < prevMB->sad8[iSubBlock]) ) ) |
if ( (iMinSAD < 256/4 ) || ( (MVequal(*currMV,prevMB->mvs[iSubBlock])) && ((uint32_t)iMinSAD < prevMB->sad8[iSubBlock]) ) ) |
1360 |
{ |
{ |
1361 |
if (MotionFlags & PMV_QUICKSTOP16) |
if (MotionFlags & PMV_QUICKSTOP16) |
1362 |
goto PMVfast8_Terminate_without_Refine; |
goto PMVfast8_Terminate_without_Refine; |
1369 |
Step 5: Calculate SAD for motion vectors taken from left block, top, top-right, and Previous frame block. |
Step 5: Calculate SAD for motion vectors taken from left block, top, top-right, and Previous frame block. |
1370 |
Also calculate (0,0) but do not subtract offset. |
Also calculate (0,0) but do not subtract offset. |
1371 |
Let MinSAD be the smallest SAD up to this point. |
Let MinSAD be the smallest SAD up to this point. |
1372 |
If MV is (0,0) subtract offset. ******** WHAT'S THIS 'OFFSET' ??? *********** |
If MV is (0,0) subtract offset. |
1373 |
*/ |
*/ |
1374 |
|
|
1375 |
// the prediction might be even better than mv16 |
// the prediction might be even better than mv16 |
1379 |
CHECK_MV8_ZERO; |
CHECK_MV8_ZERO; |
1380 |
|
|
1381 |
// previous frame MV is always possible |
// previous frame MV is always possible |
1382 |
CHECK_MV8_CANDIDATE(pMB->mvs[iSubBlock].x,pMB->mvs[iSubBlock].y); |
CHECK_MV8_CANDIDATE(prevMB->mvs[iSubBlock].x,prevMB->mvs[iSubBlock].y); |
1383 |
|
|
1384 |
// left neighbour, if allowed |
// left neighbour, if allowed |
1385 |
if (psad[1] != MV_MAX_ERROR) |
if (psad[1] != MV_MAX_ERROR) |
1411 |
} |
} |
1412 |
} |
} |
1413 |
|
|
1414 |
|
if ( (MVzero(*currMV)) && (!MVzero(pmv[0])) /* && (iMinSAD <= iQuant * 96) */ ) |
1415 |
|
iMinSAD -= MV8_00_BIAS; |
1416 |
|
|
1417 |
|
|
1418 |
/* Step 6: If MinSAD <= thresa goto Step 10. |
/* Step 6: If MinSAD <= thresa goto Step 10. |
1419 |
If Motion Vector equal to Previous frame motion vector and MinSAD<PrevFrmSAD goto Step 10. |
If Motion Vector equal to Previous frame motion vector and MinSAD<PrevFrmSAD goto Step 10. |
1420 |
*/ |
*/ |
1421 |
|
|
1422 |
if ( (iMinSAD <= threshA) || ( MVequal(*currMV,pMB->mvs[iSubBlock]) && ((uint32_t)iMinSAD < prevMB->sad8[iSubBlock]) ) ) |
if ( (iMinSAD <= threshA) || ( MVequal(*currMV,prevMB->mvs[iSubBlock]) && ((uint32_t)iMinSAD < prevMB->sad8[iSubBlock]) ) ) |
1423 |
{ |
{ |
1424 |
if (MotionFlags & PMV_QUICKSTOP16) |
if (MotionFlags & PMV_QUICKSTOP16) |
1425 |
goto PMVfast8_Terminate_without_Refine; |
goto PMVfast8_Terminate_without_Refine; |
1549 |
MainSearch16FuncPtr EPZSMainSearchPtr; |
MainSearch16FuncPtr EPZSMainSearchPtr; |
1550 |
|
|
1551 |
if (oldMBs == NULL) |
if (oldMBs == NULL) |
1552 |
{ oldMBs = (MACROBLOCK*) calloc(1,iWcount*iHcount*sizeof(MACROBLOCK)); |
{ oldMBs = (MACROBLOCK*) calloc(iWcount*iHcount,sizeof(MACROBLOCK)); |
1553 |
fprintf(stderr,"allocated %d bytes for oldMBs\n",iWcount*iHcount*sizeof(MACROBLOCK)); |
// fprintf(stderr,"allocated %d bytes for oldMBs\n",iWcount*iHcount*sizeof(MACROBLOCK)); |
1554 |
} |
} |
1555 |
oldMB = oldMBs + x + y * iWcount; |
oldMB = oldMBs + x + y * iWcount; |
1556 |
|
|
1558 |
get_range(&min_dx, &max_dx, &min_dy, &max_dy, |
get_range(&min_dx, &max_dx, &min_dy, &max_dy, |
1559 |
x, y, 16, iWidth, iHeight, iFcode); |
x, y, 16, iWidth, iHeight, iFcode); |
1560 |
|
|
|
/* we work with abs. MVs, not relative to prediction, so get_range is called relative to 0,0 */ |
|
|
|
|
1561 |
if (!(MotionFlags & PMV_HALFPEL16 )) |
if (!(MotionFlags & PMV_HALFPEL16 )) |
1562 |
{ min_dx = EVEN(min_dx); |
{ min_dx = EVEN(min_dx); |
1563 |
max_dx = EVEN(max_dx); |
max_dx = EVEN(max_dx); |
1597 |
iMinSAD = sad16( cur, |
iMinSAD = sad16( cur, |
1598 |
get_ref_mv(pRef, pRefH, pRefV, pRefHV, x, y, 16, currMV, iEdgedWidth), |
get_ref_mv(pRef, pRefH, pRefV, pRefHV, x, y, 16, currMV, iEdgedWidth), |
1599 |
iEdgedWidth, MV_MAX_ERROR); |
iEdgedWidth, MV_MAX_ERROR); |
1600 |
iMinSAD += calc_delta_16(currMV->x-pmv[0].x, currMV->y-pmv[0].y, (uint8_t)iFcode) * iQuant; |
iMinSAD += calc_delta_16(currMV->x-pmv[0].x, currMV->y-pmv[0].y, (uint8_t)iFcode, iQuant); |
1601 |
|
|
1602 |
// thresh1 is fixed to 256 |
// thresh1 is fixed to 256 |
1603 |
if ( (iMinSAD < 256 ) || ( (MVequal(*currMV,pMB->mvs[0])) && ((uint32_t)iMinSAD < prevMB->sad16) ) ) |
if ( (iMinSAD < 256 ) || ( (MVequal(*currMV, prevMB->mvs[0])) && ((uint32_t)iMinSAD < prevMB->sad16) ) ) |
1604 |
{ |
{ |
1605 |
if (MotionFlags & PMV_QUICKSTOP16) |
if (MotionFlags & PMV_QUICKSTOP16) |
1606 |
goto EPZS16_Terminate_without_Refine; |
goto EPZS16_Terminate_without_Refine; |
1611 |
/************** This is predictor SET B: (0,0), prev.frame MV, neighbours **************/ |
/************** This is predictor SET B: (0,0), prev.frame MV, neighbours **************/ |
1612 |
|
|
1613 |
// previous frame MV |
// previous frame MV |
1614 |
CHECK_MV16_CANDIDATE(pMB->mvs[0].x,pMB->mvs[0].y); |
CHECK_MV16_CANDIDATE(prevMB->mvs[0].x,prevMB->mvs[0].y); |
1615 |
|
|
1616 |
// set threshhold based on Min of Prediction and SAD of collocated block |
// set threshhold based on Min of Prediction and SAD of collocated block |
1617 |
// CHECK_MV16 always uses iSAD for the SAD of last vector to check, so now iSAD is what we want |
// CHECK_MV16 always uses iSAD for the SAD of last vector to check, so now iSAD is what we want |
1667 |
*/ |
*/ |
1668 |
|
|
1669 |
if ( (iMinSAD <= thresh2) |
if ( (iMinSAD <= thresh2) |
1670 |
|| ( MVequal(*currMV,pMB->mvs[0]) && ((uint32_t)iMinSAD <= prevMB->sad16) ) ) |
|| ( MVequal(*currMV,prevMB->mvs[0]) && ((uint32_t)iMinSAD <= prevMB->sad16) ) ) |
1671 |
{ |
{ |
1672 |
if (MotionFlags & PMV_QUICKSTOP16) |
if (MotionFlags & PMV_QUICKSTOP16) |
1673 |
goto EPZS16_Terminate_without_Refine; |
goto EPZS16_Terminate_without_Refine; |
1677 |
|
|
1678 |
/***** predictor SET C: acceleration MV (new!), neighbours in prev. frame(new!) ****/ |
/***** predictor SET C: acceleration MV (new!), neighbours in prev. frame(new!) ****/ |
1679 |
|
|
1680 |
backupMV = pMB->mvs[0]; // last MV |
backupMV = prevMB->mvs[0]; // collocated MV |
1681 |
backupMV.x += (pMB->mvs[0].x - oldMB->mvs[0].x ); // acceleration X |
backupMV.x += (prevMB->mvs[0].x - oldMB->mvs[0].x ); // acceleration X |
1682 |
backupMV.y += (pMB->mvs[0].y - oldMB->mvs[0].y ); // acceleration Y |
backupMV.y += (prevMB->mvs[0].y - oldMB->mvs[0].y ); // acceleration Y |
1683 |
|
|
1684 |
CHECK_MV16_CANDIDATE(backupMV.x,backupMV.y); |
CHECK_MV16_CANDIDATE(backupMV.x,backupMV.y); |
1685 |
|
|
1686 |
// left neighbour |
// left neighbour |
1687 |
if (x != 0) |
if (x != 0) |
1688 |
CHECK_MV16_CANDIDATE((oldMB-1)->mvs[0].x,oldMB->mvs[0].y); |
CHECK_MV16_CANDIDATE((prevMB-1)->mvs[0].x,(prevMB-1)->mvs[0].y); |
1689 |
|
|
1690 |
// top neighbour |
// top neighbour |
1691 |
if (y != 0) |
if (y != 0) |
1692 |
CHECK_MV16_CANDIDATE((oldMB-iWcount)->mvs[0].x,oldMB->mvs[0].y); |
CHECK_MV16_CANDIDATE((prevMB-iWcount)->mvs[0].x,(prevMB-iWcount)->mvs[0].y); |
1693 |
|
|
1694 |
// right neighbour, if allowed (this value is not written yet, so take it from pMB->mvs |
// right neighbour, if allowed (this value is not written yet, so take it from pMB->mvs |
1695 |
|
|
1696 |
if ((uint32_t)x != iWcount-1) |
if ((uint32_t)x != iWcount-1) |
1697 |
CHECK_MV16_CANDIDATE((pMB+1)->mvs[0].x,oldMB->mvs[0].y); |
CHECK_MV16_CANDIDATE((prevMB+1)->mvs[0].x,(prevMB+1)->mvs[0].y); |
1698 |
|
|
1699 |
// bottom neighbour, dito |
// bottom neighbour, dito |
1700 |
if ((uint32_t)y != iHcount-1) |
if ((uint32_t)y != iHcount-1) |
1701 |
CHECK_MV16_CANDIDATE((pMB+iWcount)->mvs[0].x,oldMB->mvs[0].y); |
CHECK_MV16_CANDIDATE((prevMB+iWcount)->mvs[0].x,(prevMB+iWcount)->mvs[0].y); |
1702 |
|
|
1703 |
/* Terminate if MinSAD <= T_3 (here T_3 = T_2) */ |
/* Terminate if MinSAD <= T_3 (here T_3 = T_2) */ |
1704 |
if (iMinSAD <= thresh2) |
if (iMinSAD <= thresh2) |
1755 |
iSAD = (*EPZSMainSearchPtr)(pRef, pRefH, pRefV, pRefHV, cur, |
iSAD = (*EPZSMainSearchPtr)(pRef, pRefH, pRefV, pRefHV, cur, |
1756 |
x, y, |
x, y, |
1757 |
0, 0, iMinSAD, &newMV, |
0, 0, iMinSAD, &newMV, |
1758 |
pmv, min_dx, max_dx, min_dy, max_dy, iEdgedWidth, /*iDiamondSize*/ 2, iFcode, iQuant, 0); |
pmv, min_dx, max_dx, min_dy, max_dy, iEdgedWidth, 2, iFcode, iQuant, 0); |
1759 |
|
|
1760 |
if (iSAD < iMinSAD) |
if (iSAD < iMinSAD) |
1761 |
{ |
{ |
1776 |
|
|
1777 |
EPZS16_Terminate_without_Refine: |
EPZS16_Terminate_without_Refine: |
1778 |
|
|
1779 |
*oldMB = *pMB; |
*oldMB = *prevMB; |
1780 |
|
|
1781 |
currPMV->x = currMV->x - pmv[0].x; |
currPMV->x = currMV->x - pmv[0].x; |
1782 |
currPMV->y = currMV->y - pmv[0].y; |
currPMV->y = currMV->y - pmv[0].y; |
1801 |
VECTOR * const currMV, |
VECTOR * const currMV, |
1802 |
VECTOR * const currPMV) |
VECTOR * const currPMV) |
1803 |
{ |
{ |
1804 |
|
/* Please not that EPZS might not be a good choice for 8x8-block motion search ! */ |
1805 |
|
|
1806 |
const uint32_t iWcount = pParam->mb_width; |
const uint32_t iWcount = pParam->mb_width; |
1807 |
const int32_t iWidth = pParam->width; |
const int32_t iWidth = pParam->width; |
1808 |
const int32_t iHeight = pParam->height; |
const int32_t iHeight = pParam->height; |
1880 |
iMinSAD = sad8( cur, |
iMinSAD = sad8( cur, |
1881 |
get_ref_mv(pRef, pRefH, pRefV, pRefHV, x, y, 8, currMV, iEdgedWidth), |
get_ref_mv(pRef, pRefH, pRefV, pRefHV, x, y, 8, currMV, iEdgedWidth), |
1882 |
iEdgedWidth); |
iEdgedWidth); |
1883 |
iMinSAD += calc_delta_8(currMV->x-pmv[0].x, currMV->y-pmv[0].y, (uint8_t)iFcode) * iQuant; |
iMinSAD += calc_delta_8(currMV->x-pmv[0].x, currMV->y-pmv[0].y, (uint8_t)iFcode, iQuant); |
1884 |
|
|
1885 |
|
|
1886 |
// thresh1 is fixed to 256 |
// thresh1 is fixed to 256 |
1894 |
|
|
1895 |
/************** This is predictor SET B: (0,0), prev.frame MV, neighbours **************/ |
/************** This is predictor SET B: (0,0), prev.frame MV, neighbours **************/ |
1896 |
|
|
|
// previous frame MV |
|
|
CHECK_MV8_CANDIDATE(pMB->mvs[0].x,pMB->mvs[0].y); |
|
1897 |
|
|
1898 |
// MV=(0,0) is often a good choice |
// MV=(0,0) is often a good choice |
|
|
|
1899 |
CHECK_MV8_ZERO; |
CHECK_MV8_ZERO; |
1900 |
|
|
1901 |
|
// previous frame MV |
1902 |
|
CHECK_MV8_CANDIDATE(prevMB->mvs[iSubBlock].x,prevMB->mvs[iSubBlock].y); |
1903 |
|
|
1904 |
|
// left neighbour, if allowed |
1905 |
|
if (psad[1] != MV_MAX_ERROR) |
1906 |
|
{ |
1907 |
|
if (!(MotionFlags & PMV_HALFPEL8 )) |
1908 |
|
{ pmv[1].x = EVEN(pmv[1].x); |
1909 |
|
pmv[1].y = EVEN(pmv[1].y); |
1910 |
|
} |
1911 |
|
CHECK_MV8_CANDIDATE(pmv[1].x,pmv[1].y); |
1912 |
|
} |
1913 |
|
|
1914 |
|
// top neighbour, if allowed |
1915 |
|
if (psad[2] != MV_MAX_ERROR) |
1916 |
|
{ |
1917 |
|
if (!(MotionFlags & PMV_HALFPEL8 )) |
1918 |
|
{ pmv[2].x = EVEN(pmv[2].x); |
1919 |
|
pmv[2].y = EVEN(pmv[2].y); |
1920 |
|
} |
1921 |
|
CHECK_MV8_CANDIDATE(pmv[2].x,pmv[2].y); |
1922 |
|
|
1923 |
|
// top right neighbour, if allowed |
1924 |
|
if (psad[3] != MV_MAX_ERROR) |
1925 |
|
{ |
1926 |
|
if (!(MotionFlags & PMV_HALFPEL8 )) |
1927 |
|
{ pmv[3].x = EVEN(pmv[3].x); |
1928 |
|
pmv[3].y = EVEN(pmv[3].y); |
1929 |
|
} |
1930 |
|
CHECK_MV8_CANDIDATE(pmv[3].x,pmv[3].y); |
1931 |
|
} |
1932 |
|
} |
1933 |
|
|
1934 |
|
/* // this bias is zero anyway, at the moment! |
1935 |
|
|
1936 |
|
if ( (MVzero(*currMV)) && (!MVzero(pmv[0])) ) // && (iMinSAD <= iQuant * 96) |
1937 |
|
iMinSAD -= MV8_00_BIAS; |
1938 |
|
|
1939 |
|
*/ |
1940 |
|
|
1941 |
/* Terminate if MinSAD <= T_2 |
/* Terminate if MinSAD <= T_2 |
1942 |
Terminate if MV[t] == MV[t-1] and MinSAD[t] <= MinSAD[t-1] |
Terminate if MV[t] == MV[t-1] and MinSAD[t] <= MinSAD[t-1] |
1943 |
*/ |
*/ |
1950 |
goto EPZS8_Terminate_with_Refine; |
goto EPZS8_Terminate_with_Refine; |
1951 |
} |
} |
1952 |
|
|
1953 |
/************ (if Diamond Search) **************/ |
/************ (Diamond Search) **************/ |
1954 |
|
|
1955 |
backupMV = *currMV; /* save best prediction, actually only for EXTSEARCH */ |
backupMV = *currMV; /* save best prediction, actually only for EXTSEARCH */ |
1956 |
|
|
1957 |
if (!(MotionFlags & PMV_HALFPELDIAMOND8)) |
if (!(MotionFlags & PMV_HALFPELDIAMOND8)) |
1958 |
iDiamondSize *= 2; |
iDiamondSize *= 2; |
1959 |
|
|
1960 |
/* default: use best prediction as starting point for one call of PMVfast_MainSearch */ |
/* default: use best prediction as starting point for one call of EPZS_MainSearch */ |
1961 |
|
|
1962 |
|
/* // there is no EPZS^2 for inter4v at the moment |
1963 |
|
|
1964 |
|
if (MotionFlags & PMV_USESQUARES8) |
1965 |
|
EPZSMainSearchPtr = Square8_MainSearch; |
1966 |
|
else |
1967 |
|
*/ |
1968 |
|
|
|
// if (MotionFlags & PMV_USESQUARES8) |
|
|
// EPZSMainSearchPtr = Square8_MainSearch; |
|
|
// else |
|
1969 |
EPZSMainSearchPtr = Diamond8_MainSearch; |
EPZSMainSearchPtr = Diamond8_MainSearch; |
1970 |
|
|
1971 |
iSAD = (*EPZSMainSearchPtr)(pRef, pRefH, pRefV, pRefHV, cur, |
iSAD = (*EPZSMainSearchPtr)(pRef, pRefH, pRefV, pRefHV, cur, |
1972 |
x, y, |
x, y, |
1973 |
currMV->x, currMV->y, iMinSAD, &newMV, |
currMV->x, currMV->y, iMinSAD, &newMV, |
1974 |
pmv, min_dx, max_dx, min_dy, max_dy, iEdgedWidth, |
pmv, min_dx, max_dx, min_dy, max_dy, iEdgedWidth, |
1975 |
iDiamondSize, iFcode, iQuant, 00); |
iDiamondSize, iFcode, iQuant, 0); |
1976 |
|
|
1977 |
|
|
1978 |
if (iSAD < iMinSAD) |
if (iSAD < iMinSAD) |
2039 |
// TODO: need to incorporate prediction here (eg. sad += calc_delta_16) |
// TODO: need to incorporate prediction here (eg. sad += calc_delta_16) |
2040 |
***************************************************************/ |
***************************************************************/ |
2041 |
|
|
2042 |
/* |
|
2043 |
void MotionEstimationBVOP( |
void MotionEstimationBVOP( |
2044 |
MBParam * const pParam, |
MBParam * const pParam, |
2045 |
FRAMEINFO * const frame, |
FRAMEINFO * const frame, |
2061 |
const uint32_t mb_height = pParam->mb_height; |
const uint32_t mb_height = pParam->mb_height; |
2062 |
const int32_t edged_width = pParam->edged_width; |
const int32_t edged_width = pParam->edged_width; |
2063 |
|
|
2064 |
int32_t i,j; |
uint32_t i,j; |
2065 |
|
|
2066 |
int32_t f_sad16; |
int32_t f_sad16; |
2067 |
int32_t b_sad16; |
int32_t b_sad16; |
2085 |
&& b_mb->mvs[0].x == 0 |
&& b_mb->mvs[0].x == 0 |
2086 |
&& b_mb->mvs[0].y == 0) |
&& b_mb->mvs[0].y == 0) |
2087 |
{ |
{ |
2088 |
mb->mode = MB_IGNORE; |
mb->mode = MODE_NOT_CODED; |
2089 |
mb->mvs[0].x = 0; |
mb->mvs[0].x = 0; |
2090 |
mb->mvs[0].y = 0; |
mb->mvs[0].y = 0; |
2091 |
mb->b_mvs[0].x = 0; |
mb->b_mvs[0].x = 0; |
2100 |
i, j, |
i, j, |
2101 |
frame->motion_flags, frame->quant, frame->fcode, |
frame->motion_flags, frame->quant, frame->fcode, |
2102 |
pParam, |
pParam, |
2103 |
f_mbs, |
f_mbs, f_mbs /* todo */, |
2104 |
&mb->mvs[0], &pmv_dontcare); // ignore pmv |
&mb->mvs[0], &pmv_dontcare); // ignore pmv |
2105 |
|
|
2106 |
// backward search |
// backward search |
2109 |
i, j, |
i, j, |
2110 |
frame->motion_flags, frame->quant, frame->bcode, |
frame->motion_flags, frame->quant, frame->bcode, |
2111 |
pParam, |
pParam, |
2112 |
b_mbs, |
b_mbs, b_mbs, /* todo */ |
2113 |
&mb->b_mvs[0], &pmv_dontcare); // ignore pmv |
&mb->b_mvs[0], &pmv_dontcare); // ignore pmv |
2114 |
|
|
2115 |
// interpolate search (simple, but effective) |
// interpolate search (simple, but effective) |
2129 |
if (f_sad16 < b_sad16) |
if (f_sad16 < b_sad16) |
2130 |
{ |
{ |
2131 |
best_sad = f_sad16; |
best_sad = f_sad16; |
2132 |
mb->mode = MB_FORWARD; |
mb->mode = MODE_FORWARD; |
2133 |
} |
} |
2134 |
else |
else |
2135 |
{ |
{ |
2136 |
best_sad = b_sad16; |
best_sad = b_sad16; |
2137 |
mb->mode = MB_BACKWARD; |
mb->mode = MODE_BACKWARD; |
2138 |
} |
} |
2139 |
|
|
2140 |
if (i_sad16 < best_sad) |
if (i_sad16 < best_sad) |
2141 |
{ |
{ |
2142 |
best_sad = i_sad16; |
best_sad = i_sad16; |
2143 |
mb->mode = MB_INTERPOLATE; |
mb->mode = MODE_INTERPOLATE; |
2144 |
} |
} |
2145 |
|
|
2146 |
if (d_sad16 < best_sad) |
if (d_sad16 < best_sad) |
2147 |
{ |
{ |
2148 |
best_sad = d_sad16; |
best_sad = d_sad16; |
2149 |
mb->mode = MB_DIRECT; |
mb->mode = MODE_DIRECT; |
2150 |
} |
} |
2151 |
|
|
2152 |
} |
} |
2153 |
} |
} |
2154 |
} |
} |
|
|
|
|
*/ |
|