node-native-win-utils 1.1.0 → 1.3.0

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
Files changed (174) hide show
  1. package/README.md +144 -27
  2. package/binding.gyp +18 -5
  3. package/dist/index.d.ts +146 -4
  4. package/dist/index.js +107 -3
  5. package/include/opencv2/core/affine.hpp +678 -0
  6. package/include/opencv2/core/async.hpp +105 -0
  7. package/include/opencv2/core/base.hpp +664 -0
  8. package/include/opencv2/core/bindings_utils.hpp +325 -0
  9. package/include/opencv2/core/bufferpool.hpp +40 -0
  10. package/include/opencv2/core/check.hpp +170 -0
  11. package/include/opencv2/core/core.hpp +48 -0
  12. package/include/opencv2/core/core_c.h +3128 -0
  13. package/include/opencv2/core/cuda/block.hpp +211 -0
  14. package/include/opencv2/core/cuda/border_interpolate.hpp +722 -0
  15. package/include/opencv2/core/cuda/color.hpp +309 -0
  16. package/include/opencv2/core/cuda/common.hpp +131 -0
  17. package/include/opencv2/core/cuda/datamov_utils.hpp +113 -0
  18. package/include/opencv2/core/cuda/detail/color_detail.hpp +2018 -0
  19. package/include/opencv2/core/cuda/detail/reduce.hpp +365 -0
  20. package/include/opencv2/core/cuda/detail/reduce_key_val.hpp +502 -0
  21. package/include/opencv2/core/cuda/detail/transform_detail.hpp +392 -0
  22. package/include/opencv2/core/cuda/detail/type_traits_detail.hpp +191 -0
  23. package/include/opencv2/core/cuda/detail/vec_distance_detail.hpp +121 -0
  24. package/include/opencv2/core/cuda/dynamic_smem.hpp +88 -0
  25. package/include/opencv2/core/cuda/emulation.hpp +269 -0
  26. package/include/opencv2/core/cuda/filters.hpp +293 -0
  27. package/include/opencv2/core/cuda/funcattrib.hpp +79 -0
  28. package/include/opencv2/core/cuda/functional.hpp +805 -0
  29. package/include/opencv2/core/cuda/limits.hpp +128 -0
  30. package/include/opencv2/core/cuda/reduce.hpp +209 -0
  31. package/include/opencv2/core/cuda/saturate_cast.hpp +292 -0
  32. package/include/opencv2/core/cuda/scan.hpp +258 -0
  33. package/include/opencv2/core/cuda/simd_functions.hpp +869 -0
  34. package/include/opencv2/core/cuda/transform.hpp +75 -0
  35. package/include/opencv2/core/cuda/type_traits.hpp +90 -0
  36. package/include/opencv2/core/cuda/utility.hpp +230 -0
  37. package/include/opencv2/core/cuda/vec_distance.hpp +232 -0
  38. package/include/opencv2/core/cuda/vec_math.hpp +923 -0
  39. package/include/opencv2/core/cuda/vec_traits.hpp +288 -0
  40. package/include/opencv2/core/cuda/warp.hpp +139 -0
  41. package/include/opencv2/core/cuda/warp_reduce.hpp +76 -0
  42. package/include/opencv2/core/cuda/warp_shuffle.hpp +162 -0
  43. package/include/opencv2/core/cuda.hpp +1279 -0
  44. package/include/opencv2/core/cuda.inl.hpp +763 -0
  45. package/include/opencv2/core/cuda_stream_accessor.hpp +86 -0
  46. package/include/opencv2/core/cuda_types.hpp +144 -0
  47. package/include/opencv2/core/cv_cpu_dispatch.h +381 -0
  48. package/include/opencv2/core/cv_cpu_helper.h +550 -0
  49. package/include/opencv2/core/cvdef.h +973 -0
  50. package/include/opencv2/core/cvstd.hpp +190 -0
  51. package/include/opencv2/core/cvstd.inl.hpp +197 -0
  52. package/include/opencv2/core/cvstd_wrapper.hpp +154 -0
  53. package/include/opencv2/core/detail/async_promise.hpp +71 -0
  54. package/include/opencv2/core/detail/dispatch_helper.impl.hpp +49 -0
  55. package/include/opencv2/core/detail/exception_ptr.hpp +27 -0
  56. package/include/opencv2/core/directx.hpp +184 -0
  57. package/include/opencv2/core/dualquaternion.hpp +979 -0
  58. package/include/opencv2/core/dualquaternion.inl.hpp +487 -0
  59. package/include/opencv2/core/eigen.hpp +402 -0
  60. package/include/opencv2/core/fast_math.hpp +433 -0
  61. package/include/opencv2/core/hal/hal.hpp +256 -0
  62. package/include/opencv2/core/hal/interface.h +190 -0
  63. package/include/opencv2/core/hal/intrin.hpp +939 -0
  64. package/include/opencv2/core/hal/intrin_avx.hpp +3177 -0
  65. package/include/opencv2/core/hal/intrin_avx512.hpp +3090 -0
  66. package/include/opencv2/core/hal/intrin_cpp.hpp +3321 -0
  67. package/include/opencv2/core/hal/intrin_forward.hpp +191 -0
  68. package/include/opencv2/core/hal/intrin_lasx.hpp +3236 -0
  69. package/include/opencv2/core/hal/intrin_msa.hpp +1887 -0
  70. package/include/opencv2/core/hal/intrin_neon.hpp +2610 -0
  71. package/include/opencv2/core/hal/intrin_rvv.hpp +3320 -0
  72. package/include/opencv2/core/hal/intrin_rvv071.hpp +2545 -0
  73. package/include/opencv2/core/hal/intrin_rvv_scalable.hpp +2080 -0
  74. package/include/opencv2/core/hal/intrin_sse.hpp +3467 -0
  75. package/include/opencv2/core/hal/intrin_sse_em.hpp +180 -0
  76. package/include/opencv2/core/hal/intrin_vsx.hpp +1608 -0
  77. package/include/opencv2/core/hal/intrin_wasm.hpp +2782 -0
  78. package/include/opencv2/core/hal/msa_macros.h +1558 -0
  79. package/include/opencv2/core/hal/simd_utils.impl.hpp +186 -0
  80. package/include/opencv2/core/llapi/llapi.h +102 -0
  81. package/include/opencv2/core/mat.hpp +3775 -0
  82. package/include/opencv2/core/mat.inl.hpp +3422 -0
  83. package/include/opencv2/core/matx.hpp +1536 -0
  84. package/include/opencv2/core/neon_utils.hpp +128 -0
  85. package/include/opencv2/core/ocl.hpp +917 -0
  86. package/include/opencv2/core/ocl_genbase.hpp +69 -0
  87. package/include/opencv2/core/opencl/ocl_defs.hpp +82 -0
  88. package/include/opencv2/core/opencl/opencl_info.hpp +212 -0
  89. package/include/opencv2/core/opencl/opencl_svm.hpp +81 -0
  90. package/include/opencv2/core/opencl/runtime/autogenerated/opencl_clblas.hpp +602 -0
  91. package/include/opencv2/core/opencl/runtime/autogenerated/opencl_clfft.hpp +146 -0
  92. package/include/opencv2/core/opencl/runtime/autogenerated/opencl_core.hpp +371 -0
  93. package/include/opencv2/core/opencl/runtime/autogenerated/opencl_core_wrappers.hpp +272 -0
  94. package/include/opencv2/core/opencl/runtime/autogenerated/opencl_gl.hpp +62 -0
  95. package/include/opencv2/core/opencl/runtime/autogenerated/opencl_gl_wrappers.hpp +42 -0
  96. package/include/opencv2/core/opencl/runtime/opencl_clblas.hpp +53 -0
  97. package/include/opencv2/core/opencl/runtime/opencl_clfft.hpp +53 -0
  98. package/include/opencv2/core/opencl/runtime/opencl_core.hpp +84 -0
  99. package/include/opencv2/core/opencl/runtime/opencl_core_wrappers.hpp +47 -0
  100. package/include/opencv2/core/opencl/runtime/opencl_gl.hpp +53 -0
  101. package/include/opencv2/core/opencl/runtime/opencl_gl_wrappers.hpp +47 -0
  102. package/include/opencv2/core/opencl/runtime/opencl_svm_20.hpp +48 -0
  103. package/include/opencv2/core/opencl/runtime/opencl_svm_definitions.hpp +42 -0
  104. package/include/opencv2/core/opencl/runtime/opencl_svm_hsa_extension.hpp +166 -0
  105. package/include/opencv2/core/opengl.hpp +733 -0
  106. package/include/opencv2/core/openvx/ovx_defs.hpp +48 -0
  107. package/include/opencv2/core/operations.hpp +610 -0
  108. package/include/opencv2/core/optim.hpp +302 -0
  109. package/include/opencv2/core/ovx.hpp +28 -0
  110. package/include/opencv2/core/parallel/backend/parallel_for.openmp.hpp +72 -0
  111. package/include/opencv2/core/parallel/backend/parallel_for.tbb.hpp +153 -0
  112. package/include/opencv2/core/parallel/parallel_backend.hpp +90 -0
  113. package/include/opencv2/core/persistence.hpp +1350 -0
  114. package/include/opencv2/core/private/cv_cpu_include_simd_declarations.hpp +30 -0
  115. package/include/opencv2/core/private.cuda.hpp +169 -0
  116. package/include/opencv2/core/private.hpp +896 -0
  117. package/include/opencv2/core/quaternion.hpp +1696 -0
  118. package/include/opencv2/core/quaternion.inl.hpp +1063 -0
  119. package/include/opencv2/core/saturate.hpp +180 -0
  120. package/include/opencv2/core/simd_intrinsics.hpp +87 -0
  121. package/include/opencv2/core/softfloat.hpp +514 -0
  122. package/include/opencv2/core/sse_utils.hpp +652 -0
  123. package/include/opencv2/core/traits.hpp +417 -0
  124. package/include/opencv2/core/types.hpp +2457 -0
  125. package/include/opencv2/core/types_c.h +2126 -0
  126. package/include/opencv2/core/utility.hpp +1229 -0
  127. package/include/opencv2/core/utils/allocator_stats.hpp +29 -0
  128. package/include/opencv2/core/utils/allocator_stats.impl.hpp +158 -0
  129. package/include/opencv2/core/utils/buffer_area.private.hpp +136 -0
  130. package/include/opencv2/core/utils/configuration.private.hpp +22 -0
  131. package/include/opencv2/core/utils/filesystem.hpp +82 -0
  132. package/include/opencv2/core/utils/filesystem.private.hpp +66 -0
  133. package/include/opencv2/core/utils/fp_control.private.hpp +29 -0
  134. package/include/opencv2/core/utils/fp_control_utils.hpp +69 -0
  135. package/include/opencv2/core/utils/instrumentation.hpp +125 -0
  136. package/include/opencv2/core/utils/lock.private.hpp +119 -0
  137. package/include/opencv2/core/utils/logger.defines.hpp +42 -0
  138. package/include/opencv2/core/utils/logger.hpp +218 -0
  139. package/include/opencv2/core/utils/logtag.hpp +28 -0
  140. package/include/opencv2/core/utils/plugin_loader.private.hpp +165 -0
  141. package/include/opencv2/core/utils/tls.hpp +235 -0
  142. package/include/opencv2/core/utils/trace.hpp +252 -0
  143. package/include/opencv2/core/utils/trace.private.hpp +421 -0
  144. package/include/opencv2/core/va_intel.hpp +75 -0
  145. package/include/opencv2/core/version.hpp +26 -0
  146. package/include/opencv2/core/vsx_utils.hpp +1047 -0
  147. package/include/opencv2/core.hpp +3365 -0
  148. package/include/opencv2/imgcodecs/imgcodecs.hpp +48 -0
  149. package/include/opencv2/imgcodecs/imgcodecs_c.h +1 -0
  150. package/include/opencv2/imgcodecs/ios.h +59 -0
  151. package/include/opencv2/imgcodecs/legacy/constants_c.h +54 -0
  152. package/include/opencv2/imgcodecs/macosx.h +20 -0
  153. package/include/opencv2/imgcodecs.hpp +407 -0
  154. package/include/opencv2/imgproc/bindings.hpp +34 -0
  155. package/include/opencv2/imgproc/detail/gcgraph.hpp +395 -0
  156. package/include/opencv2/imgproc/hal/hal.hpp +246 -0
  157. package/include/opencv2/imgproc/hal/interface.h +46 -0
  158. package/include/opencv2/imgproc/imgproc.hpp +48 -0
  159. package/include/opencv2/imgproc/imgproc_c.h +1177 -0
  160. package/include/opencv2/imgproc/segmentation.hpp +141 -0
  161. package/include/opencv2/imgproc/types_c.h +659 -0
  162. package/include/opencv2/imgproc.hpp +5035 -0
  163. package/include/opencv2/opencv_modules.hpp +17 -0
  164. package/libs/libjpeg-turbo.lib +0 -0
  165. package/libs/libpng.lib +0 -0
  166. package/libs/opencv_core470.lib +0 -0
  167. package/libs/opencv_imgcodecs470.lib +0 -0
  168. package/libs/opencv_imgproc470.lib +0 -0
  169. package/libs/zlib.lib +0 -0
  170. package/package.json +14 -3
  171. package/prebuilds/win32-x64/node.napi.node +0 -0
  172. package/src/cpp/capturewindow.cpp +36 -46
  173. package/src/cpp/main.cpp +10 -2
  174. package/src/cpp/opencv.cpp +425 -0
@@ -0,0 +1,392 @@
1
+ /*M///////////////////////////////////////////////////////////////////////////////////////
2
+ //
3
+ // IMPORTANT: READ BEFORE DOWNLOADING, COPYING, INSTALLING OR USING.
4
+ //
5
+ // By downloading, copying, installing or using the software you agree to this license.
6
+ // If you do not agree to this license, do not download, install,
7
+ // copy or use the software.
8
+ //
9
+ //
10
+ // License Agreement
11
+ // For Open Source Computer Vision Library
12
+ //
13
+ // Copyright (C) 2000-2008, Intel Corporation, all rights reserved.
14
+ // Copyright (C) 2009, Willow Garage Inc., all rights reserved.
15
+ // Third party copyrights are property of their respective owners.
16
+ //
17
+ // Redistribution and use in source and binary forms, with or without modification,
18
+ // are permitted provided that the following conditions are met:
19
+ //
20
+ // * Redistribution's of source code must retain the above copyright notice,
21
+ // this list of conditions and the following disclaimer.
22
+ //
23
+ // * Redistribution's in binary form must reproduce the above copyright notice,
24
+ // this list of conditions and the following disclaimer in the documentation
25
+ // and/or other materials provided with the distribution.
26
+ //
27
+ // * The name of the copyright holders may not be used to endorse or promote products
28
+ // derived from this software without specific prior written permission.
29
+ //
30
+ // This software is provided by the copyright holders and contributors "as is" and
31
+ // any express or implied warranties, including, but not limited to, the implied
32
+ // warranties of merchantability and fitness for a particular purpose are disclaimed.
33
+ // In no event shall the Intel Corporation or contributors be liable for any direct,
34
+ // indirect, incidental, special, exemplary, or consequential damages
35
+ // (including, but not limited to, procurement of substitute goods or services;
36
+ // loss of use, data, or profits; or business interruption) however caused
37
+ // and on any theory of liability, whether in contract, strict liability,
38
+ // or tort (including negligence or otherwise) arising in any way out of
39
+ // the use of this software, even if advised of the possibility of such damage.
40
+ //
41
+ //M*/
42
+
43
+ #ifndef OPENCV_CUDA_TRANSFORM_DETAIL_HPP
44
+ #define OPENCV_CUDA_TRANSFORM_DETAIL_HPP
45
+
46
+ #include "../common.hpp"
47
+ #include "../vec_traits.hpp"
48
+ #include "../functional.hpp"
49
+
50
+ //! @cond IGNORED
51
+
52
+ namespace cv { namespace cuda { namespace device
53
+ {
54
+ namespace transform_detail
55
+ {
56
+ //! Read Write Traits
57
+
58
+ template <typename T, typename D, int shift> struct UnaryReadWriteTraits
59
+ {
60
+ typedef typename TypeVec<T, shift>::vec_type read_type;
61
+ typedef typename TypeVec<D, shift>::vec_type write_type;
62
+ };
63
+
64
+ template <typename T1, typename T2, typename D, int shift> struct BinaryReadWriteTraits
65
+ {
66
+ typedef typename TypeVec<T1, shift>::vec_type read_type1;
67
+ typedef typename TypeVec<T2, shift>::vec_type read_type2;
68
+ typedef typename TypeVec<D, shift>::vec_type write_type;
69
+ };
70
+
71
+ //! Transform kernels
72
+
73
+ template <int shift> struct OpUnroller;
74
+ template <> struct OpUnroller<1>
75
+ {
76
+ template <typename T, typename D, typename UnOp, typename Mask>
77
+ static __device__ __forceinline__ void unroll(const T& src, D& dst, const Mask& mask, UnOp& op, int x_shifted, int y)
78
+ {
79
+ if (mask(y, x_shifted))
80
+ dst.x = op(src.x);
81
+ }
82
+
83
+ template <typename T1, typename T2, typename D, typename BinOp, typename Mask>
84
+ static __device__ __forceinline__ void unroll(const T1& src1, const T2& src2, D& dst, const Mask& mask, BinOp& op, int x_shifted, int y)
85
+ {
86
+ if (mask(y, x_shifted))
87
+ dst.x = op(src1.x, src2.x);
88
+ }
89
+ };
90
+ template <> struct OpUnroller<2>
91
+ {
92
+ template <typename T, typename D, typename UnOp, typename Mask>
93
+ static __device__ __forceinline__ void unroll(const T& src, D& dst, const Mask& mask, UnOp& op, int x_shifted, int y)
94
+ {
95
+ if (mask(y, x_shifted))
96
+ dst.x = op(src.x);
97
+ if (mask(y, x_shifted + 1))
98
+ dst.y = op(src.y);
99
+ }
100
+
101
+ template <typename T1, typename T2, typename D, typename BinOp, typename Mask>
102
+ static __device__ __forceinline__ void unroll(const T1& src1, const T2& src2, D& dst, const Mask& mask, BinOp& op, int x_shifted, int y)
103
+ {
104
+ if (mask(y, x_shifted))
105
+ dst.x = op(src1.x, src2.x);
106
+ if (mask(y, x_shifted + 1))
107
+ dst.y = op(src1.y, src2.y);
108
+ }
109
+ };
110
+ template <> struct OpUnroller<3>
111
+ {
112
+ template <typename T, typename D, typename UnOp, typename Mask>
113
+ static __device__ __forceinline__ void unroll(const T& src, D& dst, const Mask& mask, const UnOp& op, int x_shifted, int y)
114
+ {
115
+ if (mask(y, x_shifted))
116
+ dst.x = op(src.x);
117
+ if (mask(y, x_shifted + 1))
118
+ dst.y = op(src.y);
119
+ if (mask(y, x_shifted + 2))
120
+ dst.z = op(src.z);
121
+ }
122
+
123
+ template <typename T1, typename T2, typename D, typename BinOp, typename Mask>
124
+ static __device__ __forceinline__ void unroll(const T1& src1, const T2& src2, D& dst, const Mask& mask, const BinOp& op, int x_shifted, int y)
125
+ {
126
+ if (mask(y, x_shifted))
127
+ dst.x = op(src1.x, src2.x);
128
+ if (mask(y, x_shifted + 1))
129
+ dst.y = op(src1.y, src2.y);
130
+ if (mask(y, x_shifted + 2))
131
+ dst.z = op(src1.z, src2.z);
132
+ }
133
+ };
134
+ template <> struct OpUnroller<4>
135
+ {
136
+ template <typename T, typename D, typename UnOp, typename Mask>
137
+ static __device__ __forceinline__ void unroll(const T& src, D& dst, const Mask& mask, const UnOp& op, int x_shifted, int y)
138
+ {
139
+ if (mask(y, x_shifted))
140
+ dst.x = op(src.x);
141
+ if (mask(y, x_shifted + 1))
142
+ dst.y = op(src.y);
143
+ if (mask(y, x_shifted + 2))
144
+ dst.z = op(src.z);
145
+ if (mask(y, x_shifted + 3))
146
+ dst.w = op(src.w);
147
+ }
148
+
149
+ template <typename T1, typename T2, typename D, typename BinOp, typename Mask>
150
+ static __device__ __forceinline__ void unroll(const T1& src1, const T2& src2, D& dst, const Mask& mask, const BinOp& op, int x_shifted, int y)
151
+ {
152
+ if (mask(y, x_shifted))
153
+ dst.x = op(src1.x, src2.x);
154
+ if (mask(y, x_shifted + 1))
155
+ dst.y = op(src1.y, src2.y);
156
+ if (mask(y, x_shifted + 2))
157
+ dst.z = op(src1.z, src2.z);
158
+ if (mask(y, x_shifted + 3))
159
+ dst.w = op(src1.w, src2.w);
160
+ }
161
+ };
162
+ template <> struct OpUnroller<8>
163
+ {
164
+ template <typename T, typename D, typename UnOp, typename Mask>
165
+ static __device__ __forceinline__ void unroll(const T& src, D& dst, const Mask& mask, const UnOp& op, int x_shifted, int y)
166
+ {
167
+ if (mask(y, x_shifted))
168
+ dst.a0 = op(src.a0);
169
+ if (mask(y, x_shifted + 1))
170
+ dst.a1 = op(src.a1);
171
+ if (mask(y, x_shifted + 2))
172
+ dst.a2 = op(src.a2);
173
+ if (mask(y, x_shifted + 3))
174
+ dst.a3 = op(src.a3);
175
+ if (mask(y, x_shifted + 4))
176
+ dst.a4 = op(src.a4);
177
+ if (mask(y, x_shifted + 5))
178
+ dst.a5 = op(src.a5);
179
+ if (mask(y, x_shifted + 6))
180
+ dst.a6 = op(src.a6);
181
+ if (mask(y, x_shifted + 7))
182
+ dst.a7 = op(src.a7);
183
+ }
184
+
185
+ template <typename T1, typename T2, typename D, typename BinOp, typename Mask>
186
+ static __device__ __forceinline__ void unroll(const T1& src1, const T2& src2, D& dst, const Mask& mask, const BinOp& op, int x_shifted, int y)
187
+ {
188
+ if (mask(y, x_shifted))
189
+ dst.a0 = op(src1.a0, src2.a0);
190
+ if (mask(y, x_shifted + 1))
191
+ dst.a1 = op(src1.a1, src2.a1);
192
+ if (mask(y, x_shifted + 2))
193
+ dst.a2 = op(src1.a2, src2.a2);
194
+ if (mask(y, x_shifted + 3))
195
+ dst.a3 = op(src1.a3, src2.a3);
196
+ if (mask(y, x_shifted + 4))
197
+ dst.a4 = op(src1.a4, src2.a4);
198
+ if (mask(y, x_shifted + 5))
199
+ dst.a5 = op(src1.a5, src2.a5);
200
+ if (mask(y, x_shifted + 6))
201
+ dst.a6 = op(src1.a6, src2.a6);
202
+ if (mask(y, x_shifted + 7))
203
+ dst.a7 = op(src1.a7, src2.a7);
204
+ }
205
+ };
206
+
207
+ template <typename T, typename D, typename UnOp, typename Mask>
208
+ static __global__ void transformSmart(const PtrStepSz<T> src_, PtrStep<D> dst_, const Mask mask, const UnOp op)
209
+ {
210
+ typedef TransformFunctorTraits<UnOp> ft;
211
+ typedef typename UnaryReadWriteTraits<T, D, ft::smart_shift>::read_type read_type;
212
+ typedef typename UnaryReadWriteTraits<T, D, ft::smart_shift>::write_type write_type;
213
+
214
+ const int x = threadIdx.x + blockIdx.x * blockDim.x;
215
+ const int y = threadIdx.y + blockIdx.y * blockDim.y;
216
+ const int x_shifted = x * ft::smart_shift;
217
+
218
+ if (y < src_.rows)
219
+ {
220
+ const T* src = src_.ptr(y);
221
+ D* dst = dst_.ptr(y);
222
+
223
+ if (x_shifted + ft::smart_shift - 1 < src_.cols)
224
+ {
225
+ const read_type src_n_el = ((const read_type*)src)[x];
226
+ OpUnroller<ft::smart_shift>::unroll(src_n_el, ((write_type*)dst)[x], mask, op, x_shifted, y);
227
+ }
228
+ else
229
+ {
230
+ for (int real_x = x_shifted; real_x < src_.cols; ++real_x)
231
+ {
232
+ if (mask(y, real_x))
233
+ dst[real_x] = op(src[real_x]);
234
+ }
235
+ }
236
+ }
237
+ }
238
+
239
+ template <typename T, typename D, typename UnOp, typename Mask>
240
+ __global__ static void transformSimple(const PtrStepSz<T> src, PtrStep<D> dst, const Mask mask, const UnOp op)
241
+ {
242
+ const int x = blockDim.x * blockIdx.x + threadIdx.x;
243
+ const int y = blockDim.y * blockIdx.y + threadIdx.y;
244
+
245
+ if (x < src.cols && y < src.rows && mask(y, x))
246
+ {
247
+ dst.ptr(y)[x] = op(src.ptr(y)[x]);
248
+ }
249
+ }
250
+
251
+ template <typename T1, typename T2, typename D, typename BinOp, typename Mask>
252
+ static __global__ void transformSmart(const PtrStepSz<T1> src1_, const PtrStep<T2> src2_, PtrStep<D> dst_,
253
+ const Mask mask, const BinOp op)
254
+ {
255
+ typedef TransformFunctorTraits<BinOp> ft;
256
+ typedef typename BinaryReadWriteTraits<T1, T2, D, ft::smart_shift>::read_type1 read_type1;
257
+ typedef typename BinaryReadWriteTraits<T1, T2, D, ft::smart_shift>::read_type2 read_type2;
258
+ typedef typename BinaryReadWriteTraits<T1, T2, D, ft::smart_shift>::write_type write_type;
259
+
260
+ const int x = threadIdx.x + blockIdx.x * blockDim.x;
261
+ const int y = threadIdx.y + blockIdx.y * blockDim.y;
262
+ const int x_shifted = x * ft::smart_shift;
263
+
264
+ if (y < src1_.rows)
265
+ {
266
+ const T1* src1 = src1_.ptr(y);
267
+ const T2* src2 = src2_.ptr(y);
268
+ D* dst = dst_.ptr(y);
269
+
270
+ if (x_shifted + ft::smart_shift - 1 < src1_.cols)
271
+ {
272
+ const read_type1 src1_n_el = ((const read_type1*)src1)[x];
273
+ const read_type2 src2_n_el = ((const read_type2*)src2)[x];
274
+
275
+ OpUnroller<ft::smart_shift>::unroll(src1_n_el, src2_n_el, ((write_type*)dst)[x], mask, op, x_shifted, y);
276
+ }
277
+ else
278
+ {
279
+ for (int real_x = x_shifted; real_x < src1_.cols; ++real_x)
280
+ {
281
+ if (mask(y, real_x))
282
+ dst[real_x] = op(src1[real_x], src2[real_x]);
283
+ }
284
+ }
285
+ }
286
+ }
287
+
288
+ template <typename T1, typename T2, typename D, typename BinOp, typename Mask>
289
+ static __global__ void transformSimple(const PtrStepSz<T1> src1, const PtrStep<T2> src2, PtrStep<D> dst,
290
+ const Mask mask, const BinOp op)
291
+ {
292
+ const int x = blockDim.x * blockIdx.x + threadIdx.x;
293
+ const int y = blockDim.y * blockIdx.y + threadIdx.y;
294
+
295
+ if (x < src1.cols && y < src1.rows && mask(y, x))
296
+ {
297
+ const T1 src1_data = src1.ptr(y)[x];
298
+ const T2 src2_data = src2.ptr(y)[x];
299
+ dst.ptr(y)[x] = op(src1_data, src2_data);
300
+ }
301
+ }
302
+
303
+ template <bool UseSmart> struct TransformDispatcher;
304
+ template<> struct TransformDispatcher<false>
305
+ {
306
+ template <typename T, typename D, typename UnOp, typename Mask>
307
+ static void call(PtrStepSz<T> src, PtrStepSz<D> dst, UnOp op, Mask mask, cudaStream_t stream)
308
+ {
309
+ typedef TransformFunctorTraits<UnOp> ft;
310
+
311
+ const dim3 threads(ft::simple_block_dim_x, ft::simple_block_dim_y, 1);
312
+ const dim3 grid(divUp(src.cols, threads.x), divUp(src.rows, threads.y), 1);
313
+
314
+ transformSimple<T, D><<<grid, threads, 0, stream>>>(src, dst, mask, op);
315
+ cudaSafeCall( cudaGetLastError() );
316
+
317
+ if (stream == 0)
318
+ cudaSafeCall( cudaDeviceSynchronize() );
319
+ }
320
+
321
+ template <typename T1, typename T2, typename D, typename BinOp, typename Mask>
322
+ static void call(PtrStepSz<T1> src1, PtrStepSz<T2> src2, PtrStepSz<D> dst, BinOp op, Mask mask, cudaStream_t stream)
323
+ {
324
+ typedef TransformFunctorTraits<BinOp> ft;
325
+
326
+ const dim3 threads(ft::simple_block_dim_x, ft::simple_block_dim_y, 1);
327
+ const dim3 grid(divUp(src1.cols, threads.x), divUp(src1.rows, threads.y), 1);
328
+
329
+ transformSimple<T1, T2, D><<<grid, threads, 0, stream>>>(src1, src2, dst, mask, op);
330
+ cudaSafeCall( cudaGetLastError() );
331
+
332
+ if (stream == 0)
333
+ cudaSafeCall( cudaDeviceSynchronize() );
334
+ }
335
+ };
336
+ template<> struct TransformDispatcher<true>
337
+ {
338
+ template <typename T, typename D, typename UnOp, typename Mask>
339
+ static void call(PtrStepSz<T> src, PtrStepSz<D> dst, UnOp op, Mask mask, cudaStream_t stream)
340
+ {
341
+ typedef TransformFunctorTraits<UnOp> ft;
342
+
343
+ CV_StaticAssert(ft::smart_shift != 1, "");
344
+
345
+ if (!isAligned(src.data, ft::smart_shift * sizeof(T)) || !isAligned(src.step, ft::smart_shift * sizeof(T)) ||
346
+ !isAligned(dst.data, ft::smart_shift * sizeof(D)) || !isAligned(dst.step, ft::smart_shift * sizeof(D)))
347
+ {
348
+ TransformDispatcher<false>::call(src, dst, op, mask, stream);
349
+ return;
350
+ }
351
+
352
+ const dim3 threads(ft::smart_block_dim_x, ft::smart_block_dim_y, 1);
353
+ const dim3 grid(divUp(src.cols, threads.x * ft::smart_shift), divUp(src.rows, threads.y), 1);
354
+
355
+ transformSmart<T, D><<<grid, threads, 0, stream>>>(src, dst, mask, op);
356
+ cudaSafeCall( cudaGetLastError() );
357
+
358
+ if (stream == 0)
359
+ cudaSafeCall( cudaDeviceSynchronize() );
360
+ }
361
+
362
+ template <typename T1, typename T2, typename D, typename BinOp, typename Mask>
363
+ static void call(PtrStepSz<T1> src1, PtrStepSz<T2> src2, PtrStepSz<D> dst, BinOp op, Mask mask, cudaStream_t stream)
364
+ {
365
+ typedef TransformFunctorTraits<BinOp> ft;
366
+
367
+ CV_StaticAssert(ft::smart_shift != 1, "");
368
+
369
+ if (!isAligned(src1.data, ft::smart_shift * sizeof(T1)) || !isAligned(src1.step, ft::smart_shift * sizeof(T1)) ||
370
+ !isAligned(src2.data, ft::smart_shift * sizeof(T2)) || !isAligned(src2.step, ft::smart_shift * sizeof(T2)) ||
371
+ !isAligned(dst.data, ft::smart_shift * sizeof(D)) || !isAligned(dst.step, ft::smart_shift * sizeof(D)))
372
+ {
373
+ TransformDispatcher<false>::call(src1, src2, dst, op, mask, stream);
374
+ return;
375
+ }
376
+
377
+ const dim3 threads(ft::smart_block_dim_x, ft::smart_block_dim_y, 1);
378
+ const dim3 grid(divUp(src1.cols, threads.x * ft::smart_shift), divUp(src1.rows, threads.y), 1);
379
+
380
+ transformSmart<T1, T2, D><<<grid, threads, 0, stream>>>(src1, src2, dst, mask, op);
381
+ cudaSafeCall( cudaGetLastError() );
382
+
383
+ if (stream == 0)
384
+ cudaSafeCall( cudaDeviceSynchronize() );
385
+ }
386
+ };
387
+ } // namespace transform_detail
388
+ }}} // namespace cv { namespace cuda { namespace cudev
389
+
390
+ //! @endcond
391
+
392
+ #endif // OPENCV_CUDA_TRANSFORM_DETAIL_HPP
@@ -0,0 +1,191 @@
1
+ /*M///////////////////////////////////////////////////////////////////////////////////////
2
+ //
3
+ // IMPORTANT: READ BEFORE DOWNLOADING, COPYING, INSTALLING OR USING.
4
+ //
5
+ // By downloading, copying, installing or using the software you agree to this license.
6
+ // If you do not agree to this license, do not download, install,
7
+ // copy or use the software.
8
+ //
9
+ //
10
+ // License Agreement
11
+ // For Open Source Computer Vision Library
12
+ //
13
+ // Copyright (C) 2000-2008, Intel Corporation, all rights reserved.
14
+ // Copyright (C) 2009, Willow Garage Inc., all rights reserved.
15
+ // Third party copyrights are property of their respective owners.
16
+ //
17
+ // Redistribution and use in source and binary forms, with or without modification,
18
+ // are permitted provided that the following conditions are met:
19
+ //
20
+ // * Redistribution's of source code must retain the above copyright notice,
21
+ // this list of conditions and the following disclaimer.
22
+ //
23
+ // * Redistribution's in binary form must reproduce the above copyright notice,
24
+ // this list of conditions and the following disclaimer in the documentation
25
+ // and/or other materials provided with the distribution.
26
+ //
27
+ // * The name of the copyright holders may not be used to endorse or promote products
28
+ // derived from this software without specific prior written permission.
29
+ //
30
+ // This software is provided by the copyright holders and contributors "as is" and
31
+ // any express or implied warranties, including, but not limited to, the implied
32
+ // warranties of merchantability and fitness for a particular purpose are disclaimed.
33
+ // In no event shall the Intel Corporation or contributors be liable for any direct,
34
+ // indirect, incidental, special, exemplary, or consequential damages
35
+ // (including, but not limited to, procurement of substitute goods or services;
36
+ // loss of use, data, or profits; or business interruption) however caused
37
+ // and on any theory of liability, whether in contract, strict liability,
38
+ // or tort (including negligence or otherwise) arising in any way out of
39
+ // the use of this software, even if advised of the possibility of such damage.
40
+ //
41
+ //M*/
42
+
43
+ #ifndef OPENCV_CUDA_TYPE_TRAITS_DETAIL_HPP
44
+ #define OPENCV_CUDA_TYPE_TRAITS_DETAIL_HPP
45
+
46
+ #include "../common.hpp"
47
+ #include "../vec_traits.hpp"
48
+
49
+ //! @cond IGNORED
50
+
51
+ namespace cv { namespace cuda { namespace device
52
+ {
53
+ namespace type_traits_detail
54
+ {
55
+ template <bool, typename T1, typename T2> struct Select { typedef T1 type; };
56
+ template <typename T1, typename T2> struct Select<false, T1, T2> { typedef T2 type; };
57
+
58
+ template <typename T> struct IsSignedIntergral { enum {value = 0}; };
59
+ template <> struct IsSignedIntergral<schar> { enum {value = 1}; };
60
+ template <> struct IsSignedIntergral<char1> { enum {value = 1}; };
61
+ template <> struct IsSignedIntergral<short> { enum {value = 1}; };
62
+ template <> struct IsSignedIntergral<short1> { enum {value = 1}; };
63
+ template <> struct IsSignedIntergral<int> { enum {value = 1}; };
64
+ template <> struct IsSignedIntergral<int1> { enum {value = 1}; };
65
+
66
+ template <typename T> struct IsUnsignedIntegral { enum {value = 0}; };
67
+ template <> struct IsUnsignedIntegral<uchar> { enum {value = 1}; };
68
+ template <> struct IsUnsignedIntegral<uchar1> { enum {value = 1}; };
69
+ template <> struct IsUnsignedIntegral<ushort> { enum {value = 1}; };
70
+ template <> struct IsUnsignedIntegral<ushort1> { enum {value = 1}; };
71
+ template <> struct IsUnsignedIntegral<uint> { enum {value = 1}; };
72
+ template <> struct IsUnsignedIntegral<uint1> { enum {value = 1}; };
73
+
74
+ template <typename T> struct IsIntegral { enum {value = IsSignedIntergral<T>::value || IsUnsignedIntegral<T>::value}; };
75
+ template <> struct IsIntegral<char> { enum {value = 1}; };
76
+ template <> struct IsIntegral<bool> { enum {value = 1}; };
77
+
78
+ template <typename T> struct IsFloat { enum {value = 0}; };
79
+ template <> struct IsFloat<float> { enum {value = 1}; };
80
+ template <> struct IsFloat<double> { enum {value = 1}; };
81
+
82
+ template <typename T> struct IsVec { enum {value = 0}; };
83
+ template <> struct IsVec<uchar1> { enum {value = 1}; };
84
+ template <> struct IsVec<uchar2> { enum {value = 1}; };
85
+ template <> struct IsVec<uchar3> { enum {value = 1}; };
86
+ template <> struct IsVec<uchar4> { enum {value = 1}; };
87
+ template <> struct IsVec<uchar8> { enum {value = 1}; };
88
+ template <> struct IsVec<char1> { enum {value = 1}; };
89
+ template <> struct IsVec<char2> { enum {value = 1}; };
90
+ template <> struct IsVec<char3> { enum {value = 1}; };
91
+ template <> struct IsVec<char4> { enum {value = 1}; };
92
+ template <> struct IsVec<char8> { enum {value = 1}; };
93
+ template <> struct IsVec<ushort1> { enum {value = 1}; };
94
+ template <> struct IsVec<ushort2> { enum {value = 1}; };
95
+ template <> struct IsVec<ushort3> { enum {value = 1}; };
96
+ template <> struct IsVec<ushort4> { enum {value = 1}; };
97
+ template <> struct IsVec<ushort8> { enum {value = 1}; };
98
+ template <> struct IsVec<short1> { enum {value = 1}; };
99
+ template <> struct IsVec<short2> { enum {value = 1}; };
100
+ template <> struct IsVec<short3> { enum {value = 1}; };
101
+ template <> struct IsVec<short4> { enum {value = 1}; };
102
+ template <> struct IsVec<short8> { enum {value = 1}; };
103
+ template <> struct IsVec<uint1> { enum {value = 1}; };
104
+ template <> struct IsVec<uint2> { enum {value = 1}; };
105
+ template <> struct IsVec<uint3> { enum {value = 1}; };
106
+ template <> struct IsVec<uint4> { enum {value = 1}; };
107
+ template <> struct IsVec<uint8> { enum {value = 1}; };
108
+ template <> struct IsVec<int1> { enum {value = 1}; };
109
+ template <> struct IsVec<int2> { enum {value = 1}; };
110
+ template <> struct IsVec<int3> { enum {value = 1}; };
111
+ template <> struct IsVec<int4> { enum {value = 1}; };
112
+ template <> struct IsVec<int8> { enum {value = 1}; };
113
+ template <> struct IsVec<float1> { enum {value = 1}; };
114
+ template <> struct IsVec<float2> { enum {value = 1}; };
115
+ template <> struct IsVec<float3> { enum {value = 1}; };
116
+ template <> struct IsVec<float4> { enum {value = 1}; };
117
+ template <> struct IsVec<float8> { enum {value = 1}; };
118
+ template <> struct IsVec<double1> { enum {value = 1}; };
119
+ template <> struct IsVec<double2> { enum {value = 1}; };
120
+ template <> struct IsVec<double3> { enum {value = 1}; };
121
+ template <> struct IsVec<double4> { enum {value = 1}; };
122
+ template <> struct IsVec<double8> { enum {value = 1}; };
123
+
124
+ template <class U> struct AddParameterType { typedef const U& type; };
125
+ template <class U> struct AddParameterType<U&> { typedef U& type; };
126
+ template <> struct AddParameterType<void> { typedef void type; };
127
+
128
+ template <class U> struct ReferenceTraits
129
+ {
130
+ enum { value = false };
131
+ typedef U type;
132
+ };
133
+ template <class U> struct ReferenceTraits<U&>
134
+ {
135
+ enum { value = true };
136
+ typedef U type;
137
+ };
138
+
139
+ template <class U> struct PointerTraits
140
+ {
141
+ enum { value = false };
142
+ typedef void type;
143
+ };
144
+ template <class U> struct PointerTraits<U*>
145
+ {
146
+ enum { value = true };
147
+ typedef U type;
148
+ };
149
+ template <class U> struct PointerTraits<U*&>
150
+ {
151
+ enum { value = true };
152
+ typedef U type;
153
+ };
154
+
155
+ template <class U> struct UnConst
156
+ {
157
+ typedef U type;
158
+ enum { value = 0 };
159
+ };
160
+ template <class U> struct UnConst<const U>
161
+ {
162
+ typedef U type;
163
+ enum { value = 1 };
164
+ };
165
+ template <class U> struct UnConst<const U&>
166
+ {
167
+ typedef U& type;
168
+ enum { value = 1 };
169
+ };
170
+
171
+ template <class U> struct UnVolatile
172
+ {
173
+ typedef U type;
174
+ enum { value = 0 };
175
+ };
176
+ template <class U> struct UnVolatile<volatile U>
177
+ {
178
+ typedef U type;
179
+ enum { value = 1 };
180
+ };
181
+ template <class U> struct UnVolatile<volatile U&>
182
+ {
183
+ typedef U& type;
184
+ enum { value = 1 };
185
+ };
186
+ } // namespace type_traits_detail
187
+ }}} // namespace cv { namespace cuda { namespace cudev
188
+
189
+ //! @endcond
190
+
191
+ #endif // OPENCV_CUDA_TYPE_TRAITS_DETAIL_HPP