forked from opencv/opencv
-
Notifications
You must be signed in to change notification settings - Fork 0
Expand file tree
/
Copy pathelement_operations.cpp
More file actions
5034 lines (4478 loc) · 205 KB
/
element_operations.cpp
File metadata and controls
5034 lines (4478 loc) · 205 KB
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
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
/*M///////////////////////////////////////////////////////////////////////////////////////
//
// IMPORTANT: READ BEFORE DOWNLOADING, COPYING, INSTALLING OR USING.
//
// By downloading, copying, installing or using the software you agree to this license.
// If you do not agree to this license, do not download, install,
// copy or use the software.
//
//
// License Agreement
// For Open Source Computer Vision Library
//
// Copyright (C) 2000-2008, Intel Corporation, all rights reserved.
// Copyright (C) 2009, Willow Garage Inc., all rights reserved.
// Third party copyrights are property of their respective owners.
//
// Redistribution and use in source and binary forms, with or without modification,
// are permitted provided that the following conditions are met:
//
// * Redistribution's of source code must retain the above copyright notice,
// this list of conditions and the following disclaimer.
//
// * Redistribution's in binary form must reproduce the above copyright notice,
// this list of conditions and the following disclaimer in the documentation
// and/or other materials provided with the distribution.
//
// * The name of the copyright holders may not be used to endorse or promote products
// derived from this software without specific prior written permission.
//
// This software is provided by the copyright holders and contributors "as is" and
// any express or implied warranties, including, but not limited to, the implied
// warranties of merchantability and fitness for a particular purpose are disclaimed.
// In no event shall the Intel Corporation or contributors be liable for any direct,
// indirect, incidental, special, exemplary, or consequential damages
// (including, but not limited to, procurement of substitute goods or services;
// loss of use, data, or profits; or business interruption) however caused
// and on any theory of liability, whether in contract, strict liability,
// or tort (including negligence or otherwise) arising in any way out of
// the use of this software, even if advised of the possibility of such damage.
//
//M*/
#include "precomp.hpp"
using namespace cv;
using namespace cv::gpu;
#if !defined (HAVE_CUDA) || defined (CUDA_DISABLER)
void cv::gpu::add(const GpuMat&, const GpuMat&, GpuMat&, const GpuMat&, int, Stream&) { throw_nogpu(); }
void cv::gpu::add(const GpuMat&, const Scalar&, GpuMat&, const GpuMat&, int, Stream&) { throw_nogpu(); }
void cv::gpu::subtract(const GpuMat&, const GpuMat&, GpuMat&, const GpuMat&, int, Stream&) { throw_nogpu(); }
void cv::gpu::subtract(const GpuMat&, const Scalar&, GpuMat&, const GpuMat&, int, Stream&) { throw_nogpu(); }
void cv::gpu::multiply(const GpuMat&, const GpuMat&, GpuMat&, double, int, Stream&) { throw_nogpu(); }
void cv::gpu::multiply(const GpuMat&, const Scalar&, GpuMat&, double, int, Stream&) { throw_nogpu(); }
void cv::gpu::divide(const GpuMat&, const GpuMat&, GpuMat&, double, int, Stream&) { throw_nogpu(); }
void cv::gpu::divide(const GpuMat&, const Scalar&, GpuMat&, double, int, Stream&) { throw_nogpu(); }
void cv::gpu::divide(double, const GpuMat&, GpuMat&, int, Stream&) { throw_nogpu(); }
void cv::gpu::absdiff(const GpuMat&, const GpuMat&, GpuMat&, Stream&) { throw_nogpu(); }
void cv::gpu::absdiff(const GpuMat&, const Scalar&, GpuMat&, Stream&) { throw_nogpu(); }
void cv::gpu::abs(const GpuMat&, GpuMat&, Stream&) { throw_nogpu(); }
void cv::gpu::sqr(const GpuMat&, GpuMat&, Stream&) { throw_nogpu(); }
void cv::gpu::sqrt(const GpuMat&, GpuMat&, Stream&) { throw_nogpu(); }
void cv::gpu::exp(const GpuMat&, GpuMat&, Stream&) { throw_nogpu(); }
void cv::gpu::log(const GpuMat&, GpuMat&, Stream&) { throw_nogpu(); }
void cv::gpu::compare(const GpuMat&, const GpuMat&, GpuMat&, int, Stream&) { throw_nogpu(); }
void cv::gpu::compare(const GpuMat&, Scalar, GpuMat&, int, Stream&) { throw_nogpu(); }
void cv::gpu::bitwise_not(const GpuMat&, GpuMat&, const GpuMat&, Stream&) { throw_nogpu(); }
void cv::gpu::bitwise_or(const GpuMat&, const GpuMat&, GpuMat&, const GpuMat&, Stream&) { throw_nogpu(); }
void cv::gpu::bitwise_or(const GpuMat&, const Scalar&, GpuMat&, Stream&) { throw_nogpu(); }
void cv::gpu::bitwise_and(const GpuMat&, const GpuMat&, GpuMat&, const GpuMat&, Stream&) { throw_nogpu(); }
void cv::gpu::bitwise_and(const GpuMat&, const Scalar&, GpuMat&, Stream&) { throw_nogpu(); }
void cv::gpu::bitwise_xor(const GpuMat&, const GpuMat&, GpuMat&, const GpuMat&, Stream&) { throw_nogpu(); }
void cv::gpu::bitwise_xor(const GpuMat&, const Scalar&, GpuMat&, Stream&) { throw_nogpu(); }
void cv::gpu::rshift(const GpuMat&, Scalar_<int>, GpuMat&, Stream&) { throw_nogpu(); }
void cv::gpu::lshift(const GpuMat&, Scalar_<int>, GpuMat&, Stream&) { throw_nogpu(); }
void cv::gpu::min(const GpuMat&, const GpuMat&, GpuMat&, Stream&) { throw_nogpu(); }
void cv::gpu::min(const GpuMat&, double, GpuMat&, Stream&) { throw_nogpu(); }
void cv::gpu::max(const GpuMat&, const GpuMat&, GpuMat&, Stream&) { throw_nogpu(); }
void cv::gpu::max(const GpuMat&, double, GpuMat&, Stream&) { throw_nogpu(); }
double cv::gpu::threshold(const GpuMat&, GpuMat&, double, double, int, Stream&) {throw_nogpu(); return 0.0;}
void cv::gpu::pow(const GpuMat&, double, GpuMat&, Stream&) { throw_nogpu(); }
void cv::gpu::alphaComp(const GpuMat&, const GpuMat&, GpuMat&, int, Stream&) { throw_nogpu(); }
void cv::gpu::addWeighted(const GpuMat&, double, const GpuMat&, double, double, GpuMat&, int, Stream&) { throw_nogpu(); }
#else
////////////////////////////////////////////////////////////////////////
// Basic arithmetical operations (add subtract multiply divide)
namespace
{
template<int DEPTH> struct NppTypeTraits;
template<> struct NppTypeTraits<CV_8U> { typedef Npp8u npp_t; };
template<> struct NppTypeTraits<CV_8S> { typedef Npp8s npp_t; };
template<> struct NppTypeTraits<CV_16U> { typedef Npp16u npp_t; };
template<> struct NppTypeTraits<CV_16S> { typedef Npp16s npp_t; typedef Npp16sc npp_complex_type; };
template<> struct NppTypeTraits<CV_32S> { typedef Npp32s npp_t; typedef Npp32sc npp_complex_type; };
template<> struct NppTypeTraits<CV_32F> { typedef Npp32f npp_t; typedef Npp32fc npp_complex_type; };
template<> struct NppTypeTraits<CV_64F> { typedef Npp64f npp_t; typedef Npp64fc npp_complex_type; };
template<int DEPTH, int cn> struct NppArithmScalarFunc
{
typedef typename NppTypeTraits<DEPTH>::npp_t npp_t;
typedef NppStatus (*func_ptr)(const npp_t* pSrc1, int nSrc1Step, const npp_t* pConstants,
npp_t* pDst, int nDstStep, NppiSize oSizeROI, int nScaleFactor);
};
template<int DEPTH> struct NppArithmScalarFunc<DEPTH, 1>
{
typedef typename NppTypeTraits<DEPTH>::npp_t npp_t;
typedef NppStatus (*func_ptr)(const npp_t* pSrc1, int nSrc1Step, const npp_t pConstants,
npp_t* pDst, int nDstStep, NppiSize oSizeROI, int nScaleFactor);
};
template<int DEPTH> struct NppArithmScalarFunc<DEPTH, 2>
{
typedef typename NppTypeTraits<DEPTH>::npp_complex_type npp_complex_type;
typedef NppStatus (*func_ptr)(const npp_complex_type* pSrc1, int nSrc1Step, const npp_complex_type pConstants,
npp_complex_type* pDst, int nDstStep, NppiSize oSizeROI, int nScaleFactor);
};
template<int cn> struct NppArithmScalarFunc<CV_32F, cn>
{
typedef NppStatus (*func_ptr)(const Npp32f* pSrc1, int nSrc1Step, const Npp32f* pConstants, Npp32f* pDst, int nDstStep, NppiSize oSizeROI);
};
template<> struct NppArithmScalarFunc<CV_32F, 1>
{
typedef NppStatus (*func_ptr)(const Npp32f* pSrc1, int nSrc1Step, const Npp32f pConstants, Npp32f* pDst, int nDstStep, NppiSize oSizeROI);
};
template<> struct NppArithmScalarFunc<CV_32F, 2>
{
typedef NppStatus (*func_ptr)(const Npp32fc* pSrc1, int nSrc1Step, const Npp32fc pConstants, Npp32fc* pDst, int nDstStep, NppiSize oSizeROI);
};
template<int DEPTH, int cn, typename NppArithmScalarFunc<DEPTH, cn>::func_ptr func> struct NppArithmScalar
{
typedef typename NppTypeTraits<DEPTH>::npp_t npp_t;
static void call(const PtrStepSzb src, Scalar sc, PtrStepb dst, cudaStream_t stream)
{
NppStreamHandler h(stream);
NppiSize sz;
sz.width = src.cols;
sz.height = src.rows;
const npp_t pConstants[] = { saturate_cast<npp_t>(sc.val[0]), saturate_cast<npp_t>(sc.val[1]), saturate_cast<npp_t>(sc.val[2]), saturate_cast<npp_t>(sc.val[3]) };
nppSafeCall( func((const npp_t*)src.data, static_cast<int>(src.step), pConstants, (npp_t*)dst.data, static_cast<int>(dst.step), sz, 0) );
if (stream == 0)
cudaSafeCall( cudaDeviceSynchronize() );
}
};
template<int DEPTH, typename NppArithmScalarFunc<DEPTH, 1>::func_ptr func> struct NppArithmScalar<DEPTH, 1, func>
{
typedef typename NppTypeTraits<DEPTH>::npp_t npp_t;
static void call(const PtrStepSzb src, Scalar sc, PtrStepb dst, cudaStream_t stream)
{
NppStreamHandler h(stream);
NppiSize sz;
sz.width = src.cols;
sz.height = src.rows;
nppSafeCall( func((const npp_t*)src.data, static_cast<int>(src.step), saturate_cast<npp_t>(sc.val[0]), (npp_t*)dst.data, static_cast<int>(dst.step), sz, 0) );
if (stream == 0)
cudaSafeCall( cudaDeviceSynchronize() );
}
};
template<int DEPTH, typename NppArithmScalarFunc<DEPTH, 2>::func_ptr func> struct NppArithmScalar<DEPTH, 2, func>
{
typedef typename NppTypeTraits<DEPTH>::npp_t npp_t;
typedef typename NppTypeTraits<DEPTH>::npp_complex_type npp_complex_type;
static void call(const PtrStepSzb src, Scalar sc, PtrStepb dst, cudaStream_t stream)
{
NppStreamHandler h(stream);
NppiSize sz;
sz.width = src.cols;
sz.height = src.rows;
npp_complex_type nConstant;
nConstant.re = saturate_cast<npp_t>(sc.val[0]);
nConstant.im = saturate_cast<npp_t>(sc.val[1]);
nppSafeCall( func((const npp_complex_type*)src.data, static_cast<int>(src.step), nConstant,
(npp_complex_type*)dst.data, static_cast<int>(dst.step), sz, 0) );
if (stream == 0)
cudaSafeCall( cudaDeviceSynchronize() );
}
};
template<int cn, typename NppArithmScalarFunc<CV_32F, cn>::func_ptr func> struct NppArithmScalar<CV_32F, cn, func>
{
typedef typename NppTypeTraits<CV_32F>::npp_t npp_t;
static void call(const PtrStepSzb src, Scalar sc, PtrStepb dst, cudaStream_t stream)
{
NppStreamHandler h(stream);
NppiSize sz;
sz.width = src.cols;
sz.height = src.rows;
const Npp32f pConstants[] = { saturate_cast<Npp32f>(sc.val[0]), saturate_cast<Npp32f>(sc.val[1]), saturate_cast<Npp32f>(sc.val[2]), saturate_cast<Npp32f>(sc.val[3]) };
nppSafeCall( func((const npp_t*)src.data, static_cast<int>(src.step), pConstants, (npp_t*)dst.data, static_cast<int>(dst.step), sz) );
if (stream == 0)
cudaSafeCall( cudaDeviceSynchronize() );
}
};
template<typename NppArithmScalarFunc<CV_32F, 1>::func_ptr func> struct NppArithmScalar<CV_32F, 1, func>
{
typedef typename NppTypeTraits<CV_32F>::npp_t npp_t;
static void call(const PtrStepSzb src, Scalar sc, PtrStepb dst, cudaStream_t stream)
{
NppStreamHandler h(stream);
NppiSize sz;
sz.width = src.cols;
sz.height = src.rows;
nppSafeCall( func((const npp_t*)src.data, static_cast<int>(src.step), saturate_cast<Npp32f>(sc.val[0]), (npp_t*)dst.data, static_cast<int>(dst.step), sz) );
if (stream == 0)
cudaSafeCall( cudaDeviceSynchronize() );
}
};
template<typename NppArithmScalarFunc<CV_32F, 2>::func_ptr func> struct NppArithmScalar<CV_32F, 2, func>
{
typedef typename NppTypeTraits<CV_32F>::npp_t npp_t;
typedef typename NppTypeTraits<CV_32F>::npp_complex_type npp_complex_type;
static void call(const PtrStepSzb src, Scalar sc, PtrStepb dst, cudaStream_t stream)
{
NppStreamHandler h(stream);
NppiSize sz;
sz.width = src.cols;
sz.height = src.rows;
Npp32fc nConstant;
nConstant.re = saturate_cast<Npp32f>(sc.val[0]);
nConstant.im = saturate_cast<Npp32f>(sc.val[1]);
nppSafeCall( func((const npp_complex_type*)src.data, static_cast<int>(src.step), nConstant, (npp_complex_type*)dst.data, static_cast<int>(dst.step), sz) );
if (stream == 0)
cudaSafeCall( cudaDeviceSynchronize() );
}
};
}
////////////////////////////////////////////////////////////////////////
// add
namespace arithm
{
void addMat_v4(PtrStepSz<unsigned int> src1, PtrStepSz<unsigned int> src2, PtrStepSz<unsigned int> dst, cudaStream_t stream);
void addMat_v2(PtrStepSz<unsigned int> src1, PtrStepSz<unsigned int> src2, PtrStepSz<unsigned int> dst, cudaStream_t stream);
template <typename T, typename D>
void addMat(PtrStepSzb src1, PtrStepSzb src2, PtrStepSzb dst, PtrStepb mask, cudaStream_t stream);
}
void cv::gpu::add(const GpuMat& src1, const GpuMat& src2, GpuMat& dst, const GpuMat& mask, int dtype, Stream& s)
{
using namespace arithm;
typedef void (*func_t)(PtrStepSzb src1, PtrStepSzb src2, PtrStepSzb dst, PtrStepb mask, cudaStream_t stream);
#ifdef OPENCV_TINY_GPU_MODULE
static const func_t funcs[7][7] =
{
{
addMat<unsigned char, unsigned char>,
0/*addMat<unsigned char, signed char>*/,
0/*addMat<unsigned char, unsigned short>*/,
0/*addMat<unsigned char, short>*/,
0/*addMat<unsigned char, int>*/,
0/*addMat<unsigned char, float>*/,
0/*addMat<unsigned char, double>*/,
},
{
0/*addMat<signed char, unsigned char>*/,
0/*addMat<signed char, signed char>*/,
0/*addMat<signed char, unsigned short>*/,
0/*addMat<signed char, short>*/,
0/*addMat<signed char, int>*/,
0/*addMat<signed char, float>*/,
0/*addMat<signed char, double>*/,
},
{
0 /*addMat<unsigned short, unsigned char>*/,
0 /*addMat<unsigned short, signed char>*/,
0/*addMat<unsigned short, unsigned short>*/,
0/*addMat<unsigned short, short>*/,
0/*addMat<unsigned short, int>*/,
0/*addMat<unsigned short, float>*/,
0/*addMat<unsigned short, double>*/,
},
{
0 /*addMat<short, unsigned char>*/,
0 /*addMat<short, signed char>*/,
0/*addMat<short, unsigned short>*/,
0/*addMat<short, short>*/,
0/*addMat<short, int>*/,
0/*addMat<short, float>*/,
0/*addMat<short, double>*/,
},
{
0 /*addMat<int, unsigned char>*/,
0 /*addMat<int, signed char>*/,
0 /*addMat<int, unsigned short>*/,
0 /*addMat<int, short>*/,
0/*addMat<int, int>*/,
0/*addMat<int, float>*/,
0/*addMat<int, double>*/,
},
{
0 /*addMat<float, unsigned char>*/,
0 /*addMat<float, signed char>*/,
0 /*addMat<float, unsigned short>*/,
0 /*addMat<float, short>*/,
0 /*addMat<float, int>*/,
addMat<float, float>,
0/*addMat<float, double>*/,
},
{
0 /*addMat<double, unsigned char>*/,
0 /*addMat<double, signed char>*/,
0 /*addMat<double, unsigned short>*/,
0 /*addMat<double, short>*/,
0 /*addMat<double, int>*/,
0 /*addMat<double, float>*/,
0/*addMat<double, double>*/,
}
};
#else
static const func_t funcs[7][7] =
{
{
addMat<unsigned char, unsigned char>,
addMat<unsigned char, signed char>,
addMat<unsigned char, unsigned short>,
addMat<unsigned char, short>,
addMat<unsigned char, int>,
addMat<unsigned char, float>,
addMat<unsigned char, double>,
},
{
addMat<signed char, unsigned char>,
addMat<signed char, signed char>,
addMat<signed char, unsigned short>,
addMat<signed char, short>,
addMat<signed char, int>,
addMat<signed char, float>,
addMat<signed char, double>,
},
{
0 /*addMat<unsigned short, unsigned char>*/,
0 /*addMat<unsigned short, signed char>*/,
addMat<unsigned short, unsigned short>,
addMat<unsigned short, short>,
addMat<unsigned short, int>,
addMat<unsigned short, float>,
addMat<unsigned short, double>,
},
{
0 /*addMat<short, unsigned char>*/,
0 /*addMat<short, signed char>*/,
addMat<short, unsigned short>,
addMat<short, short>,
addMat<short, int>,
addMat<short, float>,
addMat<short, double>,
},
{
0 /*addMat<int, unsigned char>*/,
0 /*addMat<int, signed char>*/,
0 /*addMat<int, unsigned short>*/,
0 /*addMat<int, short>*/,
addMat<int, int>,
addMat<int, float>,
addMat<int, double>,
},
{
0 /*addMat<float, unsigned char>*/,
0 /*addMat<float, signed char>*/,
0 /*addMat<float, unsigned short>*/,
0 /*addMat<float, short>*/,
0 /*addMat<float, int>*/,
addMat<float, float>,
addMat<float, double>,
},
{
0 /*addMat<double, unsigned char>*/,
0 /*addMat<double, signed char>*/,
0 /*addMat<double, unsigned short>*/,
0 /*addMat<double, short>*/,
0 /*addMat<double, int>*/,
0 /*addMat<double, float>*/,
addMat<double, double>,
}
};
#endif
if (dtype < 0)
dtype = src1.depth();
const int sdepth = src1.depth();
const int ddepth = CV_MAT_DEPTH(dtype);
const int cn = src1.channels();
CV_Assert( sdepth <= CV_64F && ddepth <= CV_64F );
CV_Assert( src2.type() == src1.type() && src2.size() == src1.size() );
CV_Assert( mask.empty() || (cn == 1 && mask.size() == src1.size() && mask.type() == CV_8U) );
if (sdepth == CV_64F || ddepth == CV_64F)
{
if (!deviceSupports(NATIVE_DOUBLE))
CV_Error(CV_StsUnsupportedFormat, "The device doesn't support double");
}
dst.create(src1.size(), CV_MAKE_TYPE(ddepth, cn));
cudaStream_t stream = StreamAccessor::getStream(s);
PtrStepSzb src1_(src1.rows, src1.cols * cn, src1.data, src1.step);
PtrStepSzb src2_(src1.rows, src1.cols * cn, src2.data, src2.step);
PtrStepSzb dst_(src1.rows, src1.cols * cn, dst.data, dst.step);
if (mask.empty() && (sdepth == CV_8U || sdepth == CV_16U) && ddepth == sdepth)
{
const intptr_t src1ptr = reinterpret_cast<intptr_t>(src1_.data);
const intptr_t src2ptr = reinterpret_cast<intptr_t>(src2_.data);
const intptr_t dstptr = reinterpret_cast<intptr_t>(dst_.data);
const bool isAllAligned = (src1ptr & 31) == 0 && (src2ptr & 31) == 0 && (dstptr & 31) == 0;
if (isAllAligned)
{
if (sdepth == CV_8U && (src1_.cols & 3) == 0)
{
const int vcols = src1_.cols >> 2;
addMat_v4(PtrStepSz<unsigned int>(src1_.rows, vcols, (unsigned int*) src1_.data, src1_.step),
PtrStepSz<unsigned int>(src1_.rows, vcols, (unsigned int*) src2_.data, src2_.step),
PtrStepSz<unsigned int>(src1_.rows, vcols, (unsigned int*) dst_.data, dst_.step),
stream);
return;
}
else if (sdepth == CV_16U && (src1_.cols & 1) == 0)
{
const int vcols = src1_.cols >> 1;
addMat_v2(PtrStepSz<unsigned int>(src1_.rows, vcols, (unsigned int*) src1_.data, src1_.step),
PtrStepSz<unsigned int>(src1_.rows, vcols, (unsigned int*) src2_.data, src2_.step),
PtrStepSz<unsigned int>(src1_.rows, vcols, (unsigned int*) dst_.data, dst_.step),
stream);
return;
}
}
}
const func_t func = funcs[sdepth][ddepth];
if (!func)
CV_Error(CV_StsUnsupportedFormat, "Unsupported combination of source and destination types");
func(src1_, src2_, dst_, mask, stream);
}
namespace arithm
{
template <typename T, typename S, typename D>
void addScalar(PtrStepSzb src1, double val, PtrStepSzb dst, PtrStepb mask, cudaStream_t stream);
}
void cv::gpu::add(const GpuMat& src, const Scalar& sc, GpuMat& dst, const GpuMat& mask, int dtype, Stream& s)
{
using namespace arithm;
typedef void (*func_t)(PtrStepSzb src1, double val, PtrStepSzb dst, PtrStepb mask, cudaStream_t stream);
#ifdef OPENCV_TINY_GPU_MODULE
static const func_t funcs[7][7] =
{
{
addScalar<unsigned char, float, unsigned char>,
0/*addScalar<unsigned char, float, signed char>*/,
0/*addScalar<unsigned char, float, unsigned short>*/,
0/*addScalar<unsigned char, float, short>*/,
0/*addScalar<unsigned char, float, int>*/,
0/*addScalar<unsigned char, float, float>*/,
0/*addScalar<unsigned char, double, double>*/,
},
{
0/*addScalar<signed char, float, unsigned char>*/,
0/*addScalar<signed char, float, signed char>*/,
0/*addScalar<signed char, float, unsigned short>*/,
0/*addScalar<signed char, float, short>*/,
0/*addScalar<signed char, float, int>*/,
0/*addScalar<signed char, float, float>*/,
0/*addScalar<signed char, double, double>*/,
},
{
0 /*addScalar<unsigned short, float, unsigned char>*/,
0 /*addScalar<unsigned short, float, signed char>*/,
0/*addScalar<unsigned short, float, unsigned short>*/,
0/*addScalar<unsigned short, float, short>*/,
0/*addScalar<unsigned short, float, int>*/,
0/*addScalar<unsigned short, float, float>*/,
0/*addScalar<unsigned short, double, double>*/,
},
{
0 /*addScalar<short, float, unsigned char>*/,
0 /*addScalar<short, float, signed char>*/,
0/*addScalar<short, float, unsigned short>*/,
0/*addScalar<short, float, short>*/,
0/*addScalar<short, float, int>*/,
0/*addScalar<short, float, float>*/,
0/*addScalar<short, double, double>*/,
},
{
0 /*addScalar<int, float, unsigned char>*/,
0 /*addScalar<int, float, signed char>*/,
0 /*addScalar<int, float, unsigned short>*/,
0 /*addScalar<int, float, short>*/,
0/*addScalar<int, float, int>*/,
0/*addScalar<int, float, float>*/,
0/*addScalar<int, double, double>*/,
},
{
0 /*addScalar<float, float, unsigned char>*/,
0 /*addScalar<float, float, signed char>*/,
0 /*addScalar<float, float, unsigned short>*/,
0 /*addScalar<float, float, short>*/,
0 /*addScalar<float, float, int>*/,
addScalar<float, float, float>,
0/*addScalar<float, double, double>*/,
},
{
0 /*addScalar<double, double, unsigned char>*/,
0 /*addScalar<double, double, signed char>*/,
0 /*addScalar<double, double, unsigned short>*/,
0 /*addScalar<double, double, short>*/,
0 /*addScalar<double, double, int>*/,
0 /*addScalar<double, double, float>*/,
0/*addScalar<double, double, double>*/,
}
};
#else
static const func_t funcs[7][7] =
{
{
addScalar<unsigned char, float, unsigned char>,
addScalar<unsigned char, float, signed char>,
addScalar<unsigned char, float, unsigned short>,
addScalar<unsigned char, float, short>,
addScalar<unsigned char, float, int>,
addScalar<unsigned char, float, float>,
addScalar<unsigned char, double, double>,
},
{
addScalar<signed char, float, unsigned char>,
addScalar<signed char, float, signed char>,
addScalar<signed char, float, unsigned short>,
addScalar<signed char, float, short>,
addScalar<signed char, float, int>,
addScalar<signed char, float, float>,
addScalar<signed char, double, double>,
},
{
0 /*addScalar<unsigned short, float, unsigned char>*/,
0 /*addScalar<unsigned short, float, signed char>*/,
addScalar<unsigned short, float, unsigned short>,
addScalar<unsigned short, float, short>,
addScalar<unsigned short, float, int>,
addScalar<unsigned short, float, float>,
addScalar<unsigned short, double, double>,
},
{
0 /*addScalar<short, float, unsigned char>*/,
0 /*addScalar<short, float, signed char>*/,
addScalar<short, float, unsigned short>,
addScalar<short, float, short>,
addScalar<short, float, int>,
addScalar<short, float, float>,
addScalar<short, double, double>,
},
{
0 /*addScalar<int, float, unsigned char>*/,
0 /*addScalar<int, float, signed char>*/,
0 /*addScalar<int, float, unsigned short>*/,
0 /*addScalar<int, float, short>*/,
addScalar<int, float, int>,
addScalar<int, float, float>,
addScalar<int, double, double>,
},
{
0 /*addScalar<float, float, unsigned char>*/,
0 /*addScalar<float, float, signed char>*/,
0 /*addScalar<float, float, unsigned short>*/,
0 /*addScalar<float, float, short>*/,
0 /*addScalar<float, float, int>*/,
addScalar<float, float, float>,
addScalar<float, double, double>,
},
{
0 /*addScalar<double, double, unsigned char>*/,
0 /*addScalar<double, double, signed char>*/,
0 /*addScalar<double, double, unsigned short>*/,
0 /*addScalar<double, double, short>*/,
0 /*addScalar<double, double, int>*/,
0 /*addScalar<double, double, float>*/,
addScalar<double, double, double>,
}
};
#endif
typedef void (*npp_func_t)(const PtrStepSzb src, Scalar sc, PtrStepb dst, cudaStream_t stream);
static const npp_func_t npp_funcs[7][4] =
{
{NppArithmScalar<CV_8U , 1, nppiAddC_8u_C1RSfs >::call, 0 , NppArithmScalar<CV_8U , 3, nppiAddC_8u_C3RSfs >::call, NppArithmScalar<CV_8U , 4, nppiAddC_8u_C4RSfs >::call},
{0 , 0 , 0 , 0 },
{NppArithmScalar<CV_16U, 1, nppiAddC_16u_C1RSfs>::call, 0 , NppArithmScalar<CV_16U, 3, nppiAddC_16u_C3RSfs>::call, NppArithmScalar<CV_16U, 4, nppiAddC_16u_C4RSfs>::call},
{NppArithmScalar<CV_16S, 1, nppiAddC_16s_C1RSfs>::call, NppArithmScalar<CV_16S, 2, nppiAddC_16sc_C1RSfs>::call, NppArithmScalar<CV_16S, 3, nppiAddC_16s_C3RSfs>::call, NppArithmScalar<CV_16S, 4, nppiAddC_16s_C4RSfs>::call},
{NppArithmScalar<CV_32S, 1, nppiAddC_32s_C1RSfs>::call, NppArithmScalar<CV_32S, 2, nppiAddC_32sc_C1RSfs>::call, NppArithmScalar<CV_32S, 3, nppiAddC_32s_C3RSfs>::call, 0 },
{NppArithmScalar<CV_32F, 1, nppiAddC_32f_C1R >::call, NppArithmScalar<CV_32F, 2, nppiAddC_32fc_C1R >::call, NppArithmScalar<CV_32F, 3, nppiAddC_32f_C3R >::call, NppArithmScalar<CV_32F, 4, nppiAddC_32f_C4R >::call},
{0 , 0 , 0 , 0 }
};
if (dtype < 0)
dtype = src.depth();
const int sdepth = src.depth();
const int ddepth = CV_MAT_DEPTH(dtype);
const int cn = src.channels();
CV_Assert( sdepth <= CV_64F && ddepth <= CV_64F );
CV_Assert( cn <= 4 );
CV_Assert( mask.empty() || (cn == 1 && mask.size() == src.size() && mask.type() == CV_8U) );
if (sdepth == CV_64F || ddepth == CV_64F)
{
if (!deviceSupports(NATIVE_DOUBLE))
CV_Error(CV_StsUnsupportedFormat, "The device doesn't support double");
}
dst.create(src.size(), CV_MAKE_TYPE(ddepth, cn));
cudaStream_t stream = StreamAccessor::getStream(s);
const npp_func_t npp_func = npp_funcs[sdepth][cn - 1];
if (ddepth == sdepth && cn > 1 && npp_func != 0)
{
npp_func(src, sc, dst, stream);
return;
}
CV_Assert( cn == 1 );
const func_t func = funcs[sdepth][ddepth];
if (!func)
CV_Error(CV_StsUnsupportedFormat, "Unsupported combination of source and destination types");
func(src, sc.val[0], dst, mask, stream);
}
////////////////////////////////////////////////////////////////////////
// subtract
namespace arithm
{
void subMat_v4(PtrStepSz<unsigned int> src1, PtrStepSz<unsigned int> src2, PtrStepSz<unsigned int> dst, cudaStream_t stream);
void subMat_v2(PtrStepSz<unsigned int> src1, PtrStepSz<unsigned int> src2, PtrStepSz<unsigned int> dst, cudaStream_t stream);
template <typename T, typename D>
void subMat(PtrStepSzb src1, PtrStepSzb src2, PtrStepSzb dst, PtrStepb mask, cudaStream_t stream);
}
void cv::gpu::subtract(const GpuMat& src1, const GpuMat& src2, GpuMat& dst, const GpuMat& mask, int dtype, Stream& s)
{
using namespace arithm;
typedef void (*func_t)(PtrStepSzb src1, PtrStepSzb src2, PtrStepSzb dst, PtrStepb mask, cudaStream_t stream);
#ifdef OPENCV_TINY_GPU_MODULE
static const func_t funcs[7][7] =
{
{
subMat<unsigned char, unsigned char>,
0/*subMat<unsigned char, signed char>*/,
0/*subMat<unsigned char, unsigned short>*/,
0/*subMat<unsigned char, short>*/,
0/*subMat<unsigned char, int>*/,
0/*subMat<unsigned char, float>*/,
0/*subMat<unsigned char, double>*/,
},
{
0/*subMat<signed char, unsigned char>*/,
0/*subMat<signed char, signed char>*/,
0/*subMat<signed char, unsigned short>*/,
0/*subMat<signed char, short>*/,
0/*subMat<signed char, int>*/,
0/*subMat<signed char, float>*/,
0/*subMat<signed char, double>*/,
},
{
0 /*subMat<unsigned short, unsigned char>*/,
0 /*subMat<unsigned short, signed char>*/,
0/*subMat<unsigned short, unsigned short>*/,
0/*subMat<unsigned short, short>*/,
0/*subMat<unsigned short, int>*/,
0/*subMat<unsigned short, float>*/,
0/*subMat<unsigned short, double>*/,
},
{
0 /*subMat<short, unsigned char>*/,
0 /*subMat<short, signed char>*/,
0/*subMat<short, unsigned short>*/,
0/*subMat<short, short>*/,
0/*subMat<short, int>*/,
0/*subMat<short, float>*/,
0/*subMat<short, double>*/,
},
{
0 /*subMat<int, unsigned char>*/,
0 /*subMat<int, signed char>*/,
0 /*subMat<int, unsigned short>*/,
0 /*subMat<int, short>*/,
0/*subMat<int, int>*/,
0/*subMat<int, float>*/,
0/*subMat<int, double>*/,
},
{
0 /*subMat<float, unsigned char>*/,
0 /*subMat<float, signed char>*/,
0 /*subMat<float, unsigned short>*/,
0 /*subMat<float, short>*/,
0 /*subMat<float, int>*/,
subMat<float, float>,
0/*subMat<float, double>*/,
},
{
0 /*subMat<double, unsigned char>*/,
0 /*subMat<double, signed char>*/,
0 /*subMat<double, unsigned short>*/,
0 /*subMat<double, short>*/,
0 /*subMat<double, int>*/,
0 /*subMat<double, float>*/,
0/*subMat<double, double>*/,
}
};
#else
static const func_t funcs[7][7] =
{
{
subMat<unsigned char, unsigned char>,
subMat<unsigned char, signed char>,
subMat<unsigned char, unsigned short>,
subMat<unsigned char, short>,
subMat<unsigned char, int>,
subMat<unsigned char, float>,
subMat<unsigned char, double>,
},
{
subMat<signed char, unsigned char>,
subMat<signed char, signed char>,
subMat<signed char, unsigned short>,
subMat<signed char, short>,
subMat<signed char, int>,
subMat<signed char, float>,
subMat<signed char, double>,
},
{
0 /*subMat<unsigned short, unsigned char>*/,
0 /*subMat<unsigned short, signed char>*/,
subMat<unsigned short, unsigned short>,
subMat<unsigned short, short>,
subMat<unsigned short, int>,
subMat<unsigned short, float>,
subMat<unsigned short, double>,
},
{
0 /*subMat<short, unsigned char>*/,
0 /*subMat<short, signed char>*/,
subMat<short, unsigned short>,
subMat<short, short>,
subMat<short, int>,
subMat<short, float>,
subMat<short, double>,
},
{
0 /*subMat<int, unsigned char>*/,
0 /*subMat<int, signed char>*/,
0 /*subMat<int, unsigned short>*/,
0 /*subMat<int, short>*/,
subMat<int, int>,
subMat<int, float>,
subMat<int, double>,
},
{
0 /*subMat<float, unsigned char>*/,
0 /*subMat<float, signed char>*/,
0 /*subMat<float, unsigned short>*/,
0 /*subMat<float, short>*/,
0 /*subMat<float, int>*/,
subMat<float, float>,
subMat<float, double>,
},
{
0 /*subMat<double, unsigned char>*/,
0 /*subMat<double, signed char>*/,
0 /*subMat<double, unsigned short>*/,
0 /*subMat<double, short>*/,
0 /*subMat<double, int>*/,
0 /*subMat<double, float>*/,
subMat<double, double>,
}
};
#endif
if (dtype < 0)
dtype = src1.depth();
const int sdepth = src1.depth();
const int ddepth = CV_MAT_DEPTH(dtype);
const int cn = src1.channels();
CV_Assert( sdepth <= CV_64F && ddepth <= CV_64F );
CV_Assert( src2.type() == src1.type() && src2.size() == src1.size() );
CV_Assert( mask.empty() || (cn == 1 && mask.size() == src1.size() && mask.type() == CV_8U) );
if (sdepth == CV_64F || ddepth == CV_64F)
{
if (!deviceSupports(NATIVE_DOUBLE))
CV_Error(CV_StsUnsupportedFormat, "The device doesn't support double");
}
dst.create(src1.size(), CV_MAKE_TYPE(ddepth, cn));
cudaStream_t stream = StreamAccessor::getStream(s);
PtrStepSzb src1_(src1.rows, src1.cols * cn, src1.data, src1.step);
PtrStepSzb src2_(src1.rows, src1.cols * cn, src2.data, src2.step);
PtrStepSzb dst_(src1.rows, src1.cols * cn, dst.data, dst.step);
if (mask.empty() && (sdepth == CV_8U || sdepth == CV_16U) && ddepth == sdepth)
{
const intptr_t src1ptr = reinterpret_cast<intptr_t>(src1_.data);
const intptr_t src2ptr = reinterpret_cast<intptr_t>(src2_.data);
const intptr_t dstptr = reinterpret_cast<intptr_t>(dst_.data);
const bool isAllAligned = (src1ptr & 31) == 0 && (src2ptr & 31) == 0 && (dstptr & 31) == 0;
if (isAllAligned)
{
if (sdepth == CV_8U && (src1_.cols & 3) == 0)
{
const int vcols = src1_.cols >> 2;
subMat_v4(PtrStepSz<unsigned int>(src1_.rows, vcols, (unsigned int*) src1_.data, src1_.step),
PtrStepSz<unsigned int>(src1_.rows, vcols, (unsigned int*) src2_.data, src2_.step),
PtrStepSz<unsigned int>(src1_.rows, vcols, (unsigned int*) dst_.data, dst_.step),
stream);
return;
}
else if (sdepth == CV_16U && (src1_.cols & 1) == 0)
{
const int vcols = src1_.cols >> 1;
subMat_v2(PtrStepSz<unsigned int>(src1_.rows, vcols, (unsigned int*) src1_.data, src1_.step),
PtrStepSz<unsigned int>(src1_.rows, vcols, (unsigned int*) src2_.data, src2_.step),
PtrStepSz<unsigned int>(src1_.rows, vcols, (unsigned int*) dst_.data, dst_.step),
stream);
return;
}
}
}
const func_t func = funcs[sdepth][ddepth];
if (!func)
CV_Error(CV_StsUnsupportedFormat, "Unsupported combination of source and destination types");
func(src1_, src2_, dst_, mask, stream);
}
namespace arithm
{
template <typename T, typename S, typename D>
void subScalar(PtrStepSzb src1, double val, PtrStepSzb dst, PtrStepb mask, cudaStream_t stream);
}
void cv::gpu::subtract(const GpuMat& src, const Scalar& sc, GpuMat& dst, const GpuMat& mask, int dtype, Stream& s)
{
using namespace arithm;
typedef void (*func_t)(PtrStepSzb src1, double val, PtrStepSzb dst, PtrStepb mask, cudaStream_t stream);
#ifdef OPENCV_TINY_GPU_MODULE
static const func_t funcs[7][7] =
{
{
subScalar<unsigned char, float, unsigned char>,
0/*subScalar<unsigned char, float, signed char>*/,
0/*subScalar<unsigned char, float, unsigned short>*/,
0/*subScalar<unsigned char, float, short>*/,
0/*subScalar<unsigned char, float, int>*/,
0/*subScalar<unsigned char, float, float>*/,
0/*subScalar<unsigned char, double, double>*/,
},
{
0/*subScalar<signed char, float, unsigned char>*/,
0/*subScalar<signed char, float, signed char>*/,
0/*subScalar<signed char, float, unsigned short>*/,
0/*subScalar<signed char, float, short>*/,
0/*subScalar<signed char, float, int>*/,
0/*subScalar<signed char, float, float>*/,
0/*subScalar<signed char, double, double>*/,
},
{
0 /*subScalar<unsigned short, float, unsigned char>*/,
0 /*subScalar<unsigned short, float, signed char>*/,
0/*subScalar<unsigned short, float, unsigned short>*/,
0/*subScalar<unsigned short, float, short>*/,
0/*subScalar<unsigned short, float, int>*/,
0/*subScalar<unsigned short, float, float>*/,
0/*subScalar<unsigned short, double, double>*/,
},
{
0 /*subScalar<short, float, unsigned char>*/,
0 /*subScalar<short, float, signed char>*/,
0/*subScalar<short, float, unsigned short>*/,
0/*subScalar<short, float, short>*/,
0/*subScalar<short, float, int>*/,
0/*subScalar<short, float, float>*/,
0/*subScalar<short, double, double>*/,
},
{
0 /*subScalar<int, float, unsigned char>*/,
0 /*subScalar<int, float, signed char>*/,
0 /*subScalar<int, float, unsigned short>*/,
0 /*subScalar<int, float, short>*/,
0/*subScalar<int, float, int>*/,
0/*subScalar<int, float, float>*/,
0/*subScalar<int, double, double>*/,
},
{
0 /*subScalar<float, float, unsigned char>*/,
0 /*subScalar<float, float, signed char>*/,
0 /*subScalar<float, float, unsigned short>*/,
0 /*subScalar<float, float, short>*/,
0 /*subScalar<float, float, int>*/,
subScalar<float, float, float>,
0/*subScalar<float, double, double>*/,
},
{
0 /*subScalar<double, double, unsigned char>*/,
0 /*subScalar<double, double, signed char>*/,
0 /*subScalar<double, double, unsigned short>*/,
0 /*subScalar<double, double, short>*/,
0 /*subScalar<double, double, int>*/,
0 /*subScalar<double, double, float>*/,
0/*subScalar<double, double, double>*/,
}
};
#else
static const func_t funcs[7][7] =
{
{
subScalar<unsigned char, float, unsigned char>,
subScalar<unsigned char, float, signed char>,
subScalar<unsigned char, float, unsigned short>,
subScalar<unsigned char, float, short>,
subScalar<unsigned char, float, int>,
subScalar<unsigned char, float, float>,
subScalar<unsigned char, double, double>,
},
{
subScalar<signed char, float, unsigned char>,
subScalar<signed char, float, signed char>,
subScalar<signed char, float, unsigned short>,
subScalar<signed char, float, short>,
subScalar<signed char, float, int>,
subScalar<signed char, float, float>,