1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
|
#ifdef HAVE_CONFIG_H
#include "config.h"
#endif
#include <stdio.h>
#include <stdlib.h>
#include <unistd.h>
#include "mjpeg_types.h"
#include "global.h"
#include "motion.h"
/* pointers on SAD functions */
uint32_t (*calc_SAD) (uint8_t * , uint8_t * );
uint32_t (*calc_SAD_uv) (uint8_t * , uint8_t * );
uint32_t (*calc_SAD_half) (uint8_t * , uint8_t * ,uint8_t *);
/* global denoiser structure defined in main.c and global.h */
extern struct DNSR_GLOBAL denoiser;
struct DNSR_VECTOR vector;
struct DNSR_VECTOR varray44[8];
struct DNSR_VECTOR varray22[8];
/*****************************************************************************
* generate a lowpassfiltered and subsampled copy *
* of the source image (src) at the destination *
* image location. *
* Lowpass-filtering is important, as this subsampled *
* image is used for motion estimation and the result *
* is more reliable when filtered. *
* only subsample actual data, but keeping full buffer size for simplicity *
*****************************************************************************/
void
subsample_frame (uint8_t * dst[3], uint8_t * src[3], int w, int h)
{
int x, y;
int Cw=w/SS_H;
int Ch=h/SS_V;
uint8_t *s = src[0];
uint8_t *s2 = src[0]+W;
uint8_t *d = dst[0];
/* subsample Y component */
for (y = 0; y < (h>>1); y += 1)
{
for (x = 0; x < w; x += 2)
{
*(d + (x>>1)) =
(
*(s + x ) +
*(s + x + 1) +
*(s2 + x ) +
*(s2 + x + 1) + 2 )>>2;
}
s+=2*W;
s2+=2*W;
d+=W;
}
#if 1
/* subsample U and V components */
s = src[1];
s2 = src[1]+W2;
d = dst[1];
for (y = 0; y < (Ch>>1); y += 1)
{
for (x = 0; x < Cw; x += 2)
{
*(d + (x>>1)) =
(
*(s + x ) +
*(s + x + 1) +
*(s2 + x ) +
*(s2 + x + 1) +2 )>>2;
}
s+=2*W2;
s2+=2*W2;
d+=W2;
}
s = src[2];
s2 = src[2]+W2;
d = dst[2];
for (y = 0; y < (Ch>>1); y += 1)
{
for (x = 0; x < Cw; x += 2)
{
*(d + (x>>1)) =
(
*(s + x ) +
*(s + x + 1) +
*(s2 + x ) +
*(s2 + x + 1) + 2 )>>2;
}
s+=2*W2;
s2+=2*W2;
d+=W2;
}
#endif
}
/*********************************************************************
* *
* SAD-function for Y without MMX/MMXE *
* *
*********************************************************************/
uint32_t
calc_SAD_noaccel (uint8_t * frm, uint8_t * ref)
{
uint8_t dx = 0;
uint8_t dy = 0;
int32_t Y = 0;
uint32_t d = 0;
for(dy=0;dy<8;dy++)
for(dx=0;dx<8;dx++)
{
Y=*(frm+dx+dy*denoiser.frame.w)-*(ref+dx+dy*denoiser.frame.w);
d+=(Y<0)? -Y:Y;
}
return d;
}
/*********************************************************************
* *
* SAD-function for Y with MMX *
* *
*********************************************************************/
uint32_t
calc_SAD_mmx (uint8_t * frm, uint8_t * ref)
{
static uint16_t a[4];
#ifdef HAVE_ASM_MMX
__asm__ __volatile__
(
" pxor %%mm0 , %%mm0; /* clear mm0 */\n"
" pxor %%mm7 , %%mm7; /* clear mm7 */\n"
" /* */\n"
" movl %1 , %%eax; /* load frameadress into eax */\n"
" movl %2 , %%ebx; /* load frameadress into ebx */\n"
" movl %3 , %%ecx; /* load width into ecx */\n"
" /* */\n"
".rept 8 ; /* Loop for 8 lines */\n"
" movq (%%eax), %%mm1; /* 8 Pixels from filtered frame to mm1 */\n"
" movq (%%ebx), %%mm2; /* 8 Pixels from reference frame to mm2 */\n"
" movq %%mm2 , %%mm3; /* hold a copy of mm2 in mm3 */\n"
" psubusb %%mm1 , %%mm3; /* positive differences between mm2 and mm1 */\n"
" psubusb %%mm2 , %%mm1; /* positive differences between mm1 and mm3 */\n"
" paddusb %%mm3 , %%mm1; /* mm1 now contains abs(mm1-mm2) */\n"
" movq %%mm1 , %%mm2; /* copy mm1 to mm2 */\n"
" punpcklbw %%mm7 , %%mm1; /* unpack mm1 into mm1 and mm2 */\n"
" punpckhbw %%mm7 , %%mm2; /* */\n"
" paddusw %%mm1 , %%mm0; /* add mm1 (stored in mm1 and mm2...) */\n"
" paddusw %%mm2 , %%mm0; /* to mm0 */\n"
" addl %%ecx , %%eax; /* add framewidth to frameaddress */\n"
" addl %%ecx , %%ebx; /* add framewidth to frameaddress */\n"
" .endr /* end loop */\n"
" /* */\n"
" movq %%mm0 , %0 ; /* make mm0 available to gcc ... */\n"
:"=m" (a)
:"m" (frm), "m" (ref), "m" (denoiser.frame.w)
:"%eax", "%ebx", "%ecx"
);
#endif
return (uint32_t)(a[0]+a[1]+a[2]+a[3]);
}
/*********************************************************************
* *
* SAD-function for Y with MMXE *
* *
*********************************************************************/
uint32_t
calc_SAD_mmxe (uint8_t * frm, uint8_t * ref)
{
static uint32_t a;
#ifdef HAVE_ASM_MMX
__asm__ __volatile__
(
" pxor %%mm0 , %%mm0; /* clear mm0 */\n"
" movl %1 , %%eax; /* load frameadress into eax */\n"
" movl %2 , %%ebx; /* load frameadress into ebx */\n"
" movl %3 , %%ecx; /* load width into ecx */\n"
" ; /* */\n"
" .rept 8 ; /* */\n"
" movq (%%eax), %%mm1; /* 8 Pixels from filtered frame to mm1 */\n"
" psadbw (%%ebx), %%mm1; /* 8 Pixels difference to mm1 */\n"
" paddusw %%mm1 , %%mm0; /* add result to mm0 */\n"
" addl %%ecx , %%eax; /* add framewidth to frameaddress */\n"
" addl %%ecx , %%ebx; /* add framewidth to frameaddress */\n"
" .endr ; /* */\n"
" /* */\n"
" movq %%mm0 , %0 ; /* make mm0 available to gcc ... */\n"
:"=m" (a)
:"m" (frm), "m" (ref), "m" (denoiser.frame.w)
:"%eax", "%ebx", "%ecx"
);
#endif
return a;
}
/*********************************************************************
* *
* SAD-function for UV without MMX/MMXE *
* *
*********************************************************************/
uint32_t
calc_SAD_uv_noaccel (uint8_t * frm, uint8_t * ref)
{
register uint8_t dx = 0;
register uint8_t dy = 0;
int32_t Y = 0;
uint32_t d = 0;
for(dy=0;dy<8/SS_V;dy++)
for(dx=0;dx<8/SS_H;dx++)
{
Y=*(frm+dx+dy*W2)-*(ref+dx+dy*W2);
d+=(Y<0)? -Y:Y;
}
return d;
}
/*********************************************************************
* *
* SAD-function for 4:2:0 UV with MMX *
* *
*********************************************************************/
uint32_t
calc_SAD_uv420_mmx (uint8_t * frm, uint8_t * ref)
{
static uint16_t a[4];
#ifdef HAVE_ASM_MMX
int32_t halfwidth = denoiser.frame.w / 2;
__asm__ __volatile__
(
" pxor %%mm0 , %%mm0; /* clear mm0 */\n"
" pxor %%mm7 , %%mm7; /* clear mm7 */\n"
" /* */\n"
" movl %1 , %%eax; /* load frameadress into eax */\n"
" movl %2 , %%ebx; /* load frameadress into ebx */\n"
" movl %3 , %%ecx; /* load width into ecx */\n"
" /* */\n"
".rept 4 ; /* Loop for 4 lines */\n"
" movd (%%eax), %%mm1; /* 4 Pixels from filtered frame to mm1 */\n"
" movd (%%ebx), %%mm2; /* 4 Pixels from reference frame to mm2 */\n"
" movq %%mm2 , %%mm3; /* hold a copy of mm2 in mm3 */\n"
" psubusb %%mm1 , %%mm3; /* positive differences between mm2 and mm1 */\n"
" psubusb %%mm2 , %%mm1; /* positive differences between mm1 and mm3 */\n"
" paddusb %%mm3 , %%mm1; /* mm2 now contains abs(mm1-mm2) */\n"
" movq %%mm1 , %%mm2; /* copy mm1 to mm2 */\n"
" punpcklbw %%mm7 , %%mm1; /* unpack mm1 into mm1 and mm2 */\n"
" punpckhbw %%mm7 , %%mm2; /* */\n"
" paddusw %%mm1 , %%mm2; /* add mm1 (stored in mm1 and mm2...) */\n"
" paddusw %%mm2 , %%mm0; /* to mm0 */\n"
" addl %%ecx , %%eax; /* add framewidth to frameaddress */\n"
" addl %%ecx , %%ebx; /* add framewidth to frameaddress */\n"
" .endr /* end loop */\n"
" /* */\n"
" movq %%mm0 , %0 ; /* make mm0 available to gcc ... */\n"
:"=m" (a)
:"m" (frm), "m" (ref), "m" (halfwidth)
:"%eax", "%ebx", "%ecx"
);
#endif
return (uint32_t)(a[0]+a[1]+a[2]+a[3]);
}
/*********************************************************************
* *
* SAD-function for 4:1:1 UV with MMX *
* *
*********************************************************************/
uint32_t
calc_SAD_uv411_mmx (uint8_t * frm, uint8_t * ref)
{
static uint16_t a[8];
#ifdef HAVE_ASM_MMX
__asm__ __volatile__
(
" pxor %%mm0 , %%mm0; /* clear mm0 */\n"
" pxor %%mm7 , %%mm7; /* clear mm7 */\n"
" /* */\n"
" movl %1 , %%eax; /* load frameadress into eax */\n"
" movl %2 , %%ebx; /* load frameadress into ebx */\n"
" movl %3 , %%ecx; /* load width into ecx */\n"
" /* */\n"
".rept 8 ; /* Loop for 8 lines */\n"
" movw (%%eax), %%dx; /* */\n"
" movd %%edx , %%mm1; /* 2 Pixels from filtered frame to mm1 */\n"
" movw (%%ebx), %%dx; /* */\n"
" movd %%edx , %%mm2; /* 2 Pixels from reference frame to mm2 */\n"
" movq %%mm2 , %%mm3; /* hold a copy of mm2 in mm3 */\n"
" psubusb %%mm1 , %%mm3; /* positive differences between mm2 and mm1 */\n"
" psubusb %%mm2 , %%mm1; /* positive differences between mm1 and mm3 */\n"
" paddusb %%mm3 , %%mm1; /* mm2 now contains abs(mm1-mm2) */\n"
" movq %%mm1 , %%mm2; /* copy mm1 to mm2 */\n"
" punpcklbw %%mm7 , %%mm1; /* unpack mm1 into mm1 and mm2 */\n"
" punpckhbw %%mm7 , %%mm2; /* */\n"
" paddusw %%mm1 , %%mm2; /* add mm1 (stored in mm1 and mm2...) */\n"
" paddusw %%mm2 , %%mm0; /* to mm0 */\n"
" addl %%ecx , %%eax; /* add framewidth to frameaddress */\n"
" addl %%ecx , %%ebx; /* add framewidth to frameaddress */\n"
" .endr /* end loop */\n"
" /* */\n"
" movq %%mm0 , %0 ; /* make mm0 available to gcc ... */\n"
:"=m" (a)
:"m" (frm), "m" (ref), "m" (W2)
:"%eax", "%ebx", "%ecx"
);
#endif
return (uint32_t)(a[0]+a[1]+a[2]+a[3]+a[4]+a[5]+a[6]+a[7]);
}
/*********************************************************************
* *
* SAD-function for 4:2:0 UV with MMXE *
* *
*********************************************************************/
uint32_t
calc_SAD_uv420_mmxe (uint8_t * frm, uint8_t * ref)
{
static uint32_t a;
#ifdef HAVE_ASM_MMX
int32_t halfwidth = denoiser.frame.w / 2;
__asm__ __volatile__
(
" pxor %%mm0 , %%mm0; /* clear mm0 */\n"
" movl %1 , %%eax; /* load frameadress into eax */\n"
" movl %2 , %%ebx; /* load frameadress into ebx */\n"
" movl %3 , %%ecx; /* load width into ecx */\n"
" ; /* */\n"
" .rept 4 ; /* */\n"
" movd (%%eax), %%mm1; /* 4 Pixels from filtered frame to mm1 */\n"
" movd (%%ebx), %%mm2; /* 4 Pixels from filtered frame to mm2 */\n"
" psadbw %%mm2 , %%mm1; /* 4 Pixels difference to mm1 */\n"
" paddusw %%mm1 , %%mm0; /* add result to mm0 */\n"
" addl %%ecx , %%eax; /* add framewidth to frameaddress */\n"
" addl %%ecx , %%ebx; /* add framewidth to frameaddress */\n"
" .endr ; /* */\n"
" /* */\n"
" movq %%mm0 , %0 ; /* make mm0 available to gcc ... */\n"
:"=m" (a)
:"m" (frm), "m" (ref), "m" (halfwidth)
:"%eax", "%ebx", "%ecx"
);
#endif
return a;
}
/*********************************************************************
* *
* SAD-function for 4:1:1 UV with MMXE *
* *
*********************************************************************/
uint32_t
calc_SAD_uv411_mmxe (uint8_t * frm, uint8_t * ref)
{
static uint32_t a;
#ifdef HAVE_ASM_MMX
__asm__ __volatile__
(
" pxor %%mm0 , %%mm0; /* clear mm0 */\n"
" movl %1 , %%eax; /* load frameadress into eax */\n"
" movl %2 , %%ebx; /* load frameadress into ebx */\n"
" movl %3 , %%ecx; /* load width into ecx */\n"
" ; /* */\n"
" .rept 8 ; /* */\n"
" movw (%%eax), %%dx; /* */\n"
" movd %%edx , %%mm1; /* 2 Pixels from filtered frame to mm1 */\n"
" movw (%%ebx), %%dx; /* */\n"
" movd %%edx , %%mm2; /* 2 Pixels from filtered frame to mm2 */\n"
" psadbw %%mm2 , %%mm1; /* 2 Pixels difference to mm1 */\n"
" paddusw %%mm1 , %%mm0; /* add result to mm0 */\n"
" addl %%ecx , %%eax; /* add framewidth to frameaddress */\n"
" addl %%ecx , %%ebx; /* add framewidth to frameaddress */\n"
" .endr ; /* */\n"
" /* */\n"
" movq %%mm0 , %0 ; /* make mm0 available to gcc ... */\n"
:"=m" (a)
:"m" (frm), "m" (ref), "m" (W2)
:"%eax", "%ebx", "%ecx"
);
#endif
return a;
}
/*********************************************************************
* *
* halfpel SAD-function for Y without MMX/MMXE *
* *
*********************************************************************/
uint32_t
calc_SAD_half_noaccel (uint8_t * ref, uint8_t * frm1, uint8_t * frm2)
{
uint8_t dx = 0;
uint8_t dy = 0;
int32_t Y = 0;
uint32_t d = 0;
for(dy=0;dy<8;dy++)
for(dx=0;dx<8;dx++)
{
Y=((*(frm1+dx+dy*denoiser.frame.w)+*(frm2+dx+dy*denoiser.frame.w))>>1)-*(ref+dx+dy*denoiser.frame.w);
d+=(Y<0)? -Y:Y;
}
return d;
}
/*********************************************************************
* *
* halfpel SAD-function for Y with MMX *
* *
*********************************************************************/
uint32_t
calc_SAD_half_mmx (uint8_t * ref, uint8_t * frm1, uint8_t * frm2)
{
static uint32_t a;
#ifdef HAVE_ASM_MMX
static uint32_t bit_mask[2] = {0x7f7f7f7f,0x7f7f7f7f};
__asm__ __volatile__
(
" pxor %%mm0 , %%mm0; /* clear mm0 */"
" movl %1 , %%eax; /* load frameadress into eax */"
" movl %2 , %%ebx; /* load frameadress into ebx */"
" movl %3 , %%ecx; /* load frameadress into ecx */"
" movl %4 , %%edx; /* load width into edx */"
" ; /* */"
" .rept 8 ; /* */"
" movq (%%eax), %%mm1; /* 8 Pixels from filtered frame to mm1 */"
" movq (%%ebx), %%mm2; /* 8 Pixels from filtered frame to mm2 (displaced) */"
" movq (%%ecx), %%mm3; /* reference to mm3 */"
" psrlq $1 , %%mm1; /* average source pixels */"
" psrlq $1 , %%mm2; /* shift right by one (divide by two) */"
" pand %5 , %%mm1; /* kill downshifted bits */"
" pand %5 , %%mm2; /* kill downshifted bits */"
" paddusw %%mm2 , %%mm1; /* add up ... */"
" movq %%mm3 , %%mm4; /* copy reference to mm4 */"
" psubusb %%mm1 , %%mm3; /* positive differences between mm2 and mm1 */"
" psubusb %%mm4 , %%mm1; /* positive differences between mm1 and mm3 */"
" paddusb %%mm3 , %%mm1; /* mm1 now contains abs(mm1-mm2) */"
" paddusw %%mm1 , %%mm0; /* add result to mm0 */"
" addl %%edx , %%eax; /* add framewidth to frameaddress */"
" addl %%edx , %%ebx; /* add framewidth to frameaddress */"
" addl %%edx , %%ecx; /* add framewidth to frameaddress */"
" .endr ; /* */"
" /* */"
" movq %%mm0 , %0 ; /* make mm0 available to gcc ... */"
:"=m" (a)
:"m" (frm1),"m" (frm2), "m" (ref), "m" (denoiser.frame.w), "m" (*bit_mask)
:"%eax", "%ebx", "%ecx", "%edx"
);
#endif
return a;
}
/*********************************************************************
* *
* halfpel SAD-function for Y with MMXE *
* *
*********************************************************************/
uint32_t
calc_SAD_half_mmxe (uint8_t * ref, uint8_t * frm1, uint8_t * frm2)
{
static uint32_t a;
#ifdef HAVE_ASM_MMX
__asm__ __volatile__
(
" pxor %%mm0 , %%mm0; /* clear mm0 */\n"
" movl %1 , %%eax; /* load frameadress into eax */\n"
" movl %2 , %%ebx; /* load frameadress into ebx */\n"
" movl %3 , %%ecx; /* load frameadress into ecx */\n"
" movl %4 , %%edx; /* load width into edx */\n"
" ; /* */\n"
" .rept 8 ; /* */\n"
" movq (%%eax), %%mm1; /* 8 Pixels from filtered frame to mm1 */\n"
" movq (%%ebx), %%mm2; /* 8 Pixels from filtered frame to mm2 (displaced) */\n"
" movq (%%ecx), %%mm3; /* 8 Pixels from reference frame to mm3 */\n"
" pavgb %%mm2 , %%mm1; /* average source pixels */\n"
" psadbw %%mm3 , %%mm1; /* 8 Pixels difference to mm1 */\n"
" paddusw %%mm1 , %%mm0; /* add result to mm0 */\n"
" addl %%edx , %%eax; /* add framewidth to frameaddress */\n"
" addl %%edx , %%ebx; /* add framewidth to frameaddress */\n"
" addl %%edx , %%ecx; /* add framewidth to frameaddress */\n"
" .endr ; /* */\n"
" /* */\n"
" movq %%mm0 , %0 ; /* make mm0 available to gcc ... */\n"
:"=m" (a)
:"m" (frm1),"m" (frm2), "m" (ref), "m" (denoiser.frame.w)
:"%eax", "%ebx", "%ecx", "%edx"
);
#endif
return a;
}
/*********************************************************************
* *
* Estimate Motion Vectors in 4 times subsampled frames *
* *
*********************************************************************/
void
mb_search_44 (uint16_t x, uint16_t y)
{
uint32_t best_SAD=0x00ffffff;
uint32_t SAD=0x00ffffff;
uint32_t SAD_uv=0x00ffffff;
uint8_t radius = denoiser.radius>>2; /* search radius /4 in pixels */
int32_t MB_ref_offset = denoiser.frame.w * (y>>2) + (x>>2);
int32_t MB_avg_offset;
int32_t MB_ref_offset_uv = (denoiser.frame.Cw) * (y>>2)/SS_V + (x>>2)/SS_H;
int32_t MB_avg_offset_uv;
int32_t last_uv_offset=0;
int16_t xx;
int16_t yy;
SAD = calc_SAD ( denoiser.frame.sub4ref[0]+MB_ref_offset,
denoiser.frame.sub4avg[0]+MB_ref_offset );
SAD += calc_SAD_uv ( denoiser.frame.sub4ref[1]+MB_ref_offset_uv,
denoiser.frame.sub4avg[1]+MB_ref_offset_uv );
SAD += calc_SAD_uv ( denoiser.frame.sub4ref[2]+MB_ref_offset_uv,
denoiser.frame.sub4avg[2]+MB_ref_offset_uv );
for(yy=-radius;yy<=radius;yy++)
for(xx=-radius;xx<=radius;xx++)
{
MB_avg_offset = MB_ref_offset+(xx)+(yy*denoiser.frame.w);
MB_avg_offset_uv = MB_ref_offset_uv+(xx/SS_H)+((yy/SS_V)*denoiser.frame.Cw);
SAD = calc_SAD ( denoiser.frame.sub4ref[0]+MB_ref_offset,
denoiser.frame.sub4avg[0]+MB_avg_offset );
if(MB_ref_offset_uv != last_uv_offset)
{
last_uv_offset=MB_ref_offset_uv;
SAD_uv = calc_SAD_uv ( denoiser.frame.sub4ref[1]+MB_ref_offset_uv,
denoiser.frame.sub4avg[1]+MB_avg_offset_uv );
SAD_uv += calc_SAD_uv ( denoiser.frame.sub4ref[2]+MB_ref_offset_uv,
denoiser.frame.sub4avg[2]+MB_avg_offset_uv );
}
SAD += SAD_uv;
SAD += xx*xx + yy*yy; /* favour center matches... */
if(SAD<=best_SAD)
{
best_SAD = SAD;
vector.x = xx;
vector.y = yy;
}
}
}
/*********************************************************************
* *
* Estimate Motion Vectors in 2 times subsampled frames *
* *
*********************************************************************/
void
mb_search_22 (uint16_t x, uint16_t y)
{
uint32_t best_SAD=0x00ffffff;
uint32_t SAD=0x00ffffff;
uint32_t SAD_uv=0x00ffffff;
int32_t MB_ref_offset = denoiser.frame.w * (y>>1) + (x>>1);
int32_t MB_avg_offset;
int32_t MB_ref_offset_uv = (denoiser.frame.Cw) * (y>>1)/SS_V + (x>>1)/SS_H;
int32_t MB_avg_offset_uv;
int32_t last_uv_offset=0;
int16_t xx;
int16_t yy;
int16_t vx=vector.x<<1;
int16_t vy=vector.y<<1;
/* motion-vectors from 44 can/will be wrong by +/- 3 pixels */
for(yy=-2;yy<=2;yy++)
for(xx=-2;xx<=2;xx++)
{
MB_avg_offset=MB_ref_offset+(xx+vx)+((yy+vy)*denoiser.frame.w);
/* MB_avg_offset_uv = MB_ref_offset_uv+((xx+vx)>>1)/SS_H+(((yy+vy)>>1)/SS_V)*(denoiser.frame.Cw); */
MB_avg_offset_uv = MB_ref_offset_uv+((xx+vx)/SS_H)+((yy+vy)/SS_V)*(denoiser.frame.Cw);
SAD = calc_SAD ( denoiser.frame.sub2ref[0]+MB_ref_offset,
denoiser.frame.sub2avg[0]+MB_avg_offset );
if(MB_ref_offset_uv != last_uv_offset)
{
last_uv_offset=MB_ref_offset_uv;
SAD_uv = calc_SAD_uv ( denoiser.frame.sub2ref[1]+MB_ref_offset_uv,
denoiser.frame.sub2avg[1]+MB_avg_offset_uv );
SAD_uv += calc_SAD_uv ( denoiser.frame.sub2ref[2]+MB_ref_offset_uv,
denoiser.frame.sub2avg[2]+MB_avg_offset_uv );
}
SAD += SAD_uv;
if(SAD<=best_SAD)
{
best_SAD = SAD;
varray22[2]=varray22[1];
varray22[1]=varray22[0];
varray22[0].x = xx+vx;
varray22[0].y = yy+vy;
vector.x = xx+vx;
vector.y = yy+vy;
}
}
}
/*********************************************************************
* *
* Estimate Motion Vectors in not subsampled frames *
* *
*********************************************************************/
void
mb_search_11 (uint16_t x, uint16_t y)
{
uint32_t best_SAD = 0x00ffffff;
uint32_t SAD=0x00ffffff;
uint32_t SAD_uv=0x00ffffff;
int32_t MB_ref_offset = denoiser.frame.w * (y) + (x);
int32_t MB_avg_offset;
int32_t MB_ref_offset_uv = (denoiser.frame.Cw) * y/SS_V + x/SS_H;
int32_t MB_avg_offset_uv;
int32_t last_uv_offset=0;
int16_t xx;
int16_t yy;
int16_t vx=vector.x<<1;
int16_t vy=vector.y<<1;
/* motion-vectors from 22 can/will be wrong by +/- 2 pixels */
for(yy=-2;yy<=2;yy++)
for(xx=-2;xx<=2;xx++)
{
MB_avg_offset=MB_ref_offset+(xx+vx)+((yy+vy)*denoiser.frame.w);
MB_avg_offset_uv = MB_ref_offset_uv+((xx+vx)/SS_H)+((yy+vy)/SS_V)*(denoiser.frame.Cw);
SAD = calc_SAD ( denoiser.frame.ref[0]+MB_ref_offset,
denoiser.frame.avg[0]+MB_avg_offset );
if(MB_ref_offset_uv != last_uv_offset)
{
last_uv_offset=MB_ref_offset_uv;
SAD_uv = calc_SAD_uv ( denoiser.frame.sub2ref[1]+MB_ref_offset_uv,
denoiser.frame.sub2avg[1]+MB_avg_offset_uv );
SAD_uv += calc_SAD_uv ( denoiser.frame.sub2ref[2]+MB_ref_offset_uv,
denoiser.frame.sub2avg[2]+MB_avg_offset_uv );
}
SAD += SAD_uv;
if(SAD<best_SAD)
{
best_SAD = SAD;
vector.SAD = SAD;
vector.x = xx+vx;
vector.y = yy+vy;
}
}
/* finally do a zero check against the found vector */
SAD = calc_SAD ( denoiser.frame.ref[0]+MB_ref_offset,
denoiser.frame.avg[0]+MB_ref_offset );
if(SAD<=best_SAD)
{
vector.x = 0;
vector.y = 0;
vector.SAD = SAD;
}
}
/*********************************************************************
* *
* Estimate Motion Vectors in not subsampled frames *
* *
*********************************************************************/
void
mb_search_00 (uint16_t x, uint16_t y)
{
uint32_t best_SAD = 0x00ffffff;
uint32_t SAD;
int32_t MB_ref_offset = denoiser.frame.w * (y) + (x);
int32_t MB_avg_offset1;
int32_t MB_avg_offset2;
int16_t xx;
int16_t yy;
int16_t vx=vector.x;
int16_t vy=vector.y;
MB_avg_offset1=MB_ref_offset+(vx)+((vy)*denoiser.frame.w);
for(yy=-1;yy<=1;yy++)
for(xx=-1;xx<=1;xx++)
{
MB_avg_offset2=MB_ref_offset+(vx+xx)+((vy+yy)*denoiser.frame.w);
SAD = calc_SAD_half (denoiser.frame.ref[0]+MB_ref_offset,
denoiser.frame.avg[0]+MB_avg_offset1,
denoiser.frame.avg[0]+MB_avg_offset2);
if(SAD<best_SAD)
{
best_SAD = SAD;
vector.x = xx+vx*2;
vector.y = yy+vy*2;
}
}
}
|