gpufilters module for image filtering

pull/836/head
Vladislav Vinogradov 12 years ago
parent 31c8b527c6
commit 84de6ce036
  1. 2
      modules/gpu/CMakeLists.txt
  2. 1
      modules/gpu/doc/gpu.rst
  3. 215
      modules/gpu/include/opencv2/gpu.hpp
  4. 106
      modules/gpu/src/cuda/imgproc.cu
  5. 9
      modules/gpufilters/CMakeLists.txt
  6. 0
      modules/gpufilters/doc/filtering.rst
  7. 8
      modules/gpufilters/doc/gpufilters.rst
  8. 269
      modules/gpufilters/include/opencv2/gpufilters.hpp
  9. 18
      modules/gpufilters/perf/perf_filters.cpp
  10. 47
      modules/gpufilters/perf/perf_main.cpp
  11. 43
      modules/gpufilters/perf/perf_precomp.cpp
  12. 64
      modules/gpufilters/perf/perf_precomp.hpp
  13. 2
      modules/gpufilters/src/cuda/column_filter.16sc1.cu
  14. 2
      modules/gpufilters/src/cuda/column_filter.16sc3.cu
  15. 2
      modules/gpufilters/src/cuda/column_filter.16sc4.cu
  16. 2
      modules/gpufilters/src/cuda/column_filter.16uc1.cu
  17. 2
      modules/gpufilters/src/cuda/column_filter.16uc3.cu
  18. 2
      modules/gpufilters/src/cuda/column_filter.16uc4.cu
  19. 2
      modules/gpufilters/src/cuda/column_filter.32fc1.cu
  20. 2
      modules/gpufilters/src/cuda/column_filter.32fc3.cu
  21. 2
      modules/gpufilters/src/cuda/column_filter.32fc4.cu
  22. 2
      modules/gpufilters/src/cuda/column_filter.32sc1.cu
  23. 2
      modules/gpufilters/src/cuda/column_filter.32sc3.cu
  24. 2
      modules/gpufilters/src/cuda/column_filter.32sc4.cu
  25. 2
      modules/gpufilters/src/cuda/column_filter.8uc1.cu
  26. 2
      modules/gpufilters/src/cuda/column_filter.8uc3.cu
  27. 2
      modules/gpufilters/src/cuda/column_filter.8uc4.cu
  28. 0
      modules/gpufilters/src/cuda/column_filter.hpp
  29. 158
      modules/gpufilters/src/cuda/filter2d.cu
  30. 2
      modules/gpufilters/src/cuda/row_filter.16sc1.cu
  31. 2
      modules/gpufilters/src/cuda/row_filter.16sc3.cu
  32. 2
      modules/gpufilters/src/cuda/row_filter.16sc4.cu
  33. 2
      modules/gpufilters/src/cuda/row_filter.16uc1.cu
  34. 2
      modules/gpufilters/src/cuda/row_filter.16uc3.cu
  35. 2
      modules/gpufilters/src/cuda/row_filter.16uc4.cu
  36. 2
      modules/gpufilters/src/cuda/row_filter.32fc1.cu
  37. 2
      modules/gpufilters/src/cuda/row_filter.32fc3.cu
  38. 2
      modules/gpufilters/src/cuda/row_filter.32fc4.cu
  39. 2
      modules/gpufilters/src/cuda/row_filter.32sc1.cu
  40. 2
      modules/gpufilters/src/cuda/row_filter.32sc3.cu
  41. 2
      modules/gpufilters/src/cuda/row_filter.32sc4.cu
  42. 2
      modules/gpufilters/src/cuda/row_filter.8uc1.cu
  43. 2
      modules/gpufilters/src/cuda/row_filter.8uc3.cu
  44. 2
      modules/gpufilters/src/cuda/row_filter.8uc4.cu
  45. 0
      modules/gpufilters/src/cuda/row_filter.hpp
  46. 24
      modules/gpufilters/src/filtering.cpp
  47. 43
      modules/gpufilters/src/precomp.cpp
  48. 59
      modules/gpufilters/src/precomp.hpp
  49. 18
      modules/gpufilters/test/test_filters.cpp
  50. 120
      modules/gpufilters/test/test_main.cpp
  51. 43
      modules/gpufilters/test/test_precomp.cpp
  52. 60
      modules/gpufilters/test/test_precomp.hpp
  53. 2
      modules/stitching/CMakeLists.txt
  54. 2
      modules/superres/CMakeLists.txt
  55. 1
      samples/cpp/CMakeLists.txt
  56. 2
      samples/gpu/CMakeLists.txt

@ -3,7 +3,7 @@ if(ANDROID OR IOS)
endif()
set(the_description "GPU-accelerated Computer Vision")
ocv_add_module(gpu opencv_imgproc opencv_calib3d opencv_objdetect opencv_video opencv_photo opencv_legacy opencv_gpuarithm)
ocv_add_module(gpu opencv_imgproc opencv_calib3d opencv_objdetect opencv_video opencv_photo opencv_legacy opencv_gpuarithm opencv_gpufilters)
ocv_module_include_directories("${CMAKE_CURRENT_SOURCE_DIR}/src/cuda")

@ -11,6 +11,5 @@ gpu. GPU-accelerated Computer Vision
image_processing
object_detection
feature_detection_and_description
image_filtering
camera_calibration_and_3d_reconstruction
video

@ -51,225 +51,12 @@
#include "opencv2/core/gpumat.hpp"
#include "opencv2/gpuarithm.hpp"
#include "opencv2/gpufilters.hpp"
#include "opencv2/imgproc.hpp"
#include "opencv2/objdetect.hpp"
#include "opencv2/features2d.hpp"
namespace cv { namespace gpu {
//////////////////////////////// Filter Engine ////////////////////////////////
/*!
The Base Class for 1D or Row-wise Filters
This is the base class for linear or non-linear filters that process 1D data.
In particular, such filters are used for the "horizontal" filtering parts in separable filters.
*/
class CV_EXPORTS BaseRowFilter_GPU
{
public:
BaseRowFilter_GPU(int ksize_, int anchor_) : ksize(ksize_), anchor(anchor_) {}
virtual ~BaseRowFilter_GPU() {}
virtual void operator()(const GpuMat& src, GpuMat& dst, Stream& stream = Stream::Null()) = 0;
int ksize, anchor;
};
/*!
The Base Class for Column-wise Filters
This is the base class for linear or non-linear filters that process columns of 2D arrays.
Such filters are used for the "vertical" filtering parts in separable filters.
*/
class CV_EXPORTS BaseColumnFilter_GPU
{
public:
BaseColumnFilter_GPU(int ksize_, int anchor_) : ksize(ksize_), anchor(anchor_) {}
virtual ~BaseColumnFilter_GPU() {}
virtual void operator()(const GpuMat& src, GpuMat& dst, Stream& stream = Stream::Null()) = 0;
int ksize, anchor;
};
/*!
The Base Class for Non-Separable 2D Filters.
This is the base class for linear or non-linear 2D filters.
*/
class CV_EXPORTS BaseFilter_GPU
{
public:
BaseFilter_GPU(const Size& ksize_, const Point& anchor_) : ksize(ksize_), anchor(anchor_) {}
virtual ~BaseFilter_GPU() {}
virtual void operator()(const GpuMat& src, GpuMat& dst, Stream& stream = Stream::Null()) = 0;
Size ksize;
Point anchor;
};
/*!
The Base Class for Filter Engine.
The class can be used to apply an arbitrary filtering operation to an image.
It contains all the necessary intermediate buffers.
*/
class CV_EXPORTS FilterEngine_GPU
{
public:
virtual ~FilterEngine_GPU() {}
virtual void apply(const GpuMat& src, GpuMat& dst, Rect roi = Rect(0,0,-1,-1), Stream& stream = Stream::Null()) = 0;
};
//! returns the non-separable filter engine with the specified filter
CV_EXPORTS Ptr<FilterEngine_GPU> createFilter2D_GPU(const Ptr<BaseFilter_GPU>& filter2D, int srcType, int dstType);
//! returns the separable filter engine with the specified filters
CV_EXPORTS Ptr<FilterEngine_GPU> createSeparableFilter_GPU(const Ptr<BaseRowFilter_GPU>& rowFilter,
const Ptr<BaseColumnFilter_GPU>& columnFilter, int srcType, int bufType, int dstType);
CV_EXPORTS Ptr<FilterEngine_GPU> createSeparableFilter_GPU(const Ptr<BaseRowFilter_GPU>& rowFilter,
const Ptr<BaseColumnFilter_GPU>& columnFilter, int srcType, int bufType, int dstType, GpuMat& buf);
//! returns horizontal 1D box filter
//! supports only CV_8UC1 source type and CV_32FC1 sum type
CV_EXPORTS Ptr<BaseRowFilter_GPU> getRowSumFilter_GPU(int srcType, int sumType, int ksize, int anchor = -1);
//! returns vertical 1D box filter
//! supports only CV_8UC1 sum type and CV_32FC1 dst type
CV_EXPORTS Ptr<BaseColumnFilter_GPU> getColumnSumFilter_GPU(int sumType, int dstType, int ksize, int anchor = -1);
//! returns 2D box filter
//! supports CV_8UC1 and CV_8UC4 source type, dst type must be the same as source type
CV_EXPORTS Ptr<BaseFilter_GPU> getBoxFilter_GPU(int srcType, int dstType, const Size& ksize, Point anchor = Point(-1, -1));
//! returns box filter engine
CV_EXPORTS Ptr<FilterEngine_GPU> createBoxFilter_GPU(int srcType, int dstType, const Size& ksize,
const Point& anchor = Point(-1,-1));
//! returns 2D morphological filter
//! only MORPH_ERODE and MORPH_DILATE are supported
//! supports CV_8UC1 and CV_8UC4 types
//! kernel must have CV_8UC1 type, one rows and cols == ksize.width * ksize.height
CV_EXPORTS Ptr<BaseFilter_GPU> getMorphologyFilter_GPU(int op, int type, const Mat& kernel, const Size& ksize,
Point anchor=Point(-1,-1));
//! returns morphological filter engine. Only MORPH_ERODE and MORPH_DILATE are supported.
CV_EXPORTS Ptr<FilterEngine_GPU> createMorphologyFilter_GPU(int op, int type, const Mat& kernel,
const Point& anchor = Point(-1,-1), int iterations = 1);
CV_EXPORTS Ptr<FilterEngine_GPU> createMorphologyFilter_GPU(int op, int type, const Mat& kernel, GpuMat& buf,
const Point& anchor = Point(-1,-1), int iterations = 1);
//! returns 2D filter with the specified kernel
//! supports CV_8U, CV_16U and CV_32F one and four channel image
CV_EXPORTS Ptr<BaseFilter_GPU> getLinearFilter_GPU(int srcType, int dstType, const Mat& kernel, Point anchor = Point(-1, -1), int borderType = BORDER_DEFAULT);
//! returns the non-separable linear filter engine
CV_EXPORTS Ptr<FilterEngine_GPU> createLinearFilter_GPU(int srcType, int dstType, const Mat& kernel,
Point anchor = Point(-1,-1), int borderType = BORDER_DEFAULT);
//! returns the primitive row filter with the specified kernel.
//! supports only CV_8UC1, CV_8UC4, CV_16SC1, CV_16SC2, CV_32SC1, CV_32FC1 source type.
//! there are two version of algorithm: NPP and OpenCV.
//! NPP calls when srcType == CV_8UC1 or srcType == CV_8UC4 and bufType == srcType,
//! otherwise calls OpenCV version.
//! NPP supports only BORDER_CONSTANT border type.
//! OpenCV version supports only CV_32F as buffer depth and
//! BORDER_REFLECT101, BORDER_REPLICATE and BORDER_CONSTANT border types.
CV_EXPORTS Ptr<BaseRowFilter_GPU> getLinearRowFilter_GPU(int srcType, int bufType, const Mat& rowKernel,
int anchor = -1, int borderType = BORDER_DEFAULT);
//! returns the primitive column filter with the specified kernel.
//! supports only CV_8UC1, CV_8UC4, CV_16SC1, CV_16SC2, CV_32SC1, CV_32FC1 dst type.
//! there are two version of algorithm: NPP and OpenCV.
//! NPP calls when dstType == CV_8UC1 or dstType == CV_8UC4 and bufType == dstType,
//! otherwise calls OpenCV version.
//! NPP supports only BORDER_CONSTANT border type.
//! OpenCV version supports only CV_32F as buffer depth and
//! BORDER_REFLECT101, BORDER_REPLICATE and BORDER_CONSTANT border types.
CV_EXPORTS Ptr<BaseColumnFilter_GPU> getLinearColumnFilter_GPU(int bufType, int dstType, const Mat& columnKernel,
int anchor = -1, int borderType = BORDER_DEFAULT);
//! returns the separable linear filter engine
CV_EXPORTS Ptr<FilterEngine_GPU> createSeparableLinearFilter_GPU(int srcType, int dstType, const Mat& rowKernel,
const Mat& columnKernel, const Point& anchor = Point(-1,-1), int rowBorderType = BORDER_DEFAULT,
int columnBorderType = -1);
CV_EXPORTS Ptr<FilterEngine_GPU> createSeparableLinearFilter_GPU(int srcType, int dstType, const Mat& rowKernel,
const Mat& columnKernel, GpuMat& buf, const Point& anchor = Point(-1,-1), int rowBorderType = BORDER_DEFAULT,
int columnBorderType = -1);
//! returns filter engine for the generalized Sobel operator
CV_EXPORTS Ptr<FilterEngine_GPU> createDerivFilter_GPU(int srcType, int dstType, int dx, int dy, int ksize,
int rowBorderType = BORDER_DEFAULT, int columnBorderType = -1);
CV_EXPORTS Ptr<FilterEngine_GPU> createDerivFilter_GPU(int srcType, int dstType, int dx, int dy, int ksize, GpuMat& buf,
int rowBorderType = BORDER_DEFAULT, int columnBorderType = -1);
//! returns the Gaussian filter engine
CV_EXPORTS Ptr<FilterEngine_GPU> createGaussianFilter_GPU(int type, Size ksize, double sigma1, double sigma2 = 0,
int rowBorderType = BORDER_DEFAULT, int columnBorderType = -1);
CV_EXPORTS Ptr<FilterEngine_GPU> createGaussianFilter_GPU(int type, Size ksize, GpuMat& buf, double sigma1, double sigma2 = 0,
int rowBorderType = BORDER_DEFAULT, int columnBorderType = -1);
//! returns maximum filter
CV_EXPORTS Ptr<BaseFilter_GPU> getMaxFilter_GPU(int srcType, int dstType, const Size& ksize, Point anchor = Point(-1,-1));
//! returns minimum filter
CV_EXPORTS Ptr<BaseFilter_GPU> getMinFilter_GPU(int srcType, int dstType, const Size& ksize, Point anchor = Point(-1,-1));
//! smooths the image using the normalized box filter
//! supports CV_8UC1, CV_8UC4 types
CV_EXPORTS void boxFilter(const GpuMat& src, GpuMat& dst, int ddepth, Size ksize, Point anchor = Point(-1,-1), Stream& stream = Stream::Null());
//! a synonym for normalized box filter
static inline void blur(const GpuMat& src, GpuMat& dst, Size ksize, Point anchor = Point(-1,-1), Stream& stream = Stream::Null())
{
boxFilter(src, dst, -1, ksize, anchor, stream);
}
//! erodes the image (applies the local minimum operator)
CV_EXPORTS void erode(const GpuMat& src, GpuMat& dst, const Mat& kernel, Point anchor = Point(-1, -1), int iterations = 1);
CV_EXPORTS void erode(const GpuMat& src, GpuMat& dst, const Mat& kernel, GpuMat& buf,
Point anchor = Point(-1, -1), int iterations = 1,
Stream& stream = Stream::Null());
//! dilates the image (applies the local maximum operator)
CV_EXPORTS void dilate(const GpuMat& src, GpuMat& dst, const Mat& kernel, Point anchor = Point(-1, -1), int iterations = 1);
CV_EXPORTS void dilate(const GpuMat& src, GpuMat& dst, const Mat& kernel, GpuMat& buf,
Point anchor = Point(-1, -1), int iterations = 1,
Stream& stream = Stream::Null());
//! applies an advanced morphological operation to the image
CV_EXPORTS void morphologyEx(const GpuMat& src, GpuMat& dst, int op, const Mat& kernel, Point anchor = Point(-1, -1), int iterations = 1);
CV_EXPORTS void morphologyEx(const GpuMat& src, GpuMat& dst, int op, const Mat& kernel, GpuMat& buf1, GpuMat& buf2,
Point anchor = Point(-1, -1), int iterations = 1, Stream& stream = Stream::Null());
//! applies non-separable 2D linear filter to the image
CV_EXPORTS void filter2D(const GpuMat& src, GpuMat& dst, int ddepth, const Mat& kernel, Point anchor=Point(-1,-1), int borderType = BORDER_DEFAULT, Stream& stream = Stream::Null());
//! applies separable 2D linear filter to the image
CV_EXPORTS void sepFilter2D(const GpuMat& src, GpuMat& dst, int ddepth, const Mat& kernelX, const Mat& kernelY,
Point anchor = Point(-1,-1), int rowBorderType = BORDER_DEFAULT, int columnBorderType = -1);
CV_EXPORTS void sepFilter2D(const GpuMat& src, GpuMat& dst, int ddepth, const Mat& kernelX, const Mat& kernelY, GpuMat& buf,
Point anchor = Point(-1,-1), int rowBorderType = BORDER_DEFAULT, int columnBorderType = -1,
Stream& stream = Stream::Null());
//! applies generalized Sobel operator to the image
CV_EXPORTS void Sobel(const GpuMat& src, GpuMat& dst, int ddepth, int dx, int dy, int ksize = 3, double scale = 1,
int rowBorderType = BORDER_DEFAULT, int columnBorderType = -1);
CV_EXPORTS void Sobel(const GpuMat& src, GpuMat& dst, int ddepth, int dx, int dy, GpuMat& buf, int ksize = 3, double scale = 1,
int rowBorderType = BORDER_DEFAULT, int columnBorderType = -1, Stream& stream = Stream::Null());
//! applies the vertical or horizontal Scharr operator to the image
CV_EXPORTS void Scharr(const GpuMat& src, GpuMat& dst, int ddepth, int dx, int dy, double scale = 1,
int rowBorderType = BORDER_DEFAULT, int columnBorderType = -1);
CV_EXPORTS void Scharr(const GpuMat& src, GpuMat& dst, int ddepth, int dx, int dy, GpuMat& buf, double scale = 1,
int rowBorderType = BORDER_DEFAULT, int columnBorderType = -1, Stream& stream = Stream::Null());
//! smooths the image using Gaussian filter.
CV_EXPORTS void GaussianBlur(const GpuMat& src, GpuMat& dst, Size ksize, double sigma1, double sigma2 = 0,
int rowBorderType = BORDER_DEFAULT, int columnBorderType = -1);
CV_EXPORTS void GaussianBlur(const GpuMat& src, GpuMat& dst, Size ksize, GpuMat& buf, double sigma1, double sigma2 = 0,
int rowBorderType = BORDER_DEFAULT, int columnBorderType = -1, Stream& stream = Stream::Null());
//! applies Laplacian operator to the image
//! supports only ksize = 1 and ksize = 3
CV_EXPORTS void Laplacian(const GpuMat& src, GpuMat& dst, int ddepth, int ksize = 1, double scale = 1, int borderType = BORDER_DEFAULT, Stream& stream = Stream::Null());
////////////////////////////// Image processing //////////////////////////////

@ -895,112 +895,6 @@ namespace cv { namespace gpu { namespace cudev
if (stream == 0)
cudaSafeCall(cudaDeviceSynchronize());
}
//////////////////////////////////////////////////////////////////////////
// filter2D
#define FILTER2D_MAX_KERNEL_SIZE 16
__constant__ float c_filter2DKernel[FILTER2D_MAX_KERNEL_SIZE * FILTER2D_MAX_KERNEL_SIZE];
template <class SrcT, typename D>
__global__ void filter2D(const SrcT src, PtrStepSz<D> dst, const int kWidth, const int kHeight, const int anchorX, const int anchorY)
{
typedef typename TypeVec<float, VecTraits<D>::cn>::vec_type sum_t;
const int x = blockIdx.x * blockDim.x + threadIdx.x;
const int y = blockIdx.y * blockDim.y + threadIdx.y;
if (x >= dst.cols || y >= dst.rows)
return;
sum_t res = VecTraits<sum_t>::all(0);
int kInd = 0;
for (int i = 0; i < kHeight; ++i)
{
for (int j = 0; j < kWidth; ++j)
res = res + src(y - anchorY + i, x - anchorX + j) * c_filter2DKernel[kInd++];
}
dst(y, x) = saturate_cast<D>(res);
}
template <typename T, typename D, template <typename> class Brd> struct Filter2DCaller;
#define IMPLEMENT_FILTER2D_TEX_READER(type) \
texture< type , cudaTextureType2D, cudaReadModeElementType> tex_filter2D_ ## type (0, cudaFilterModePoint, cudaAddressModeClamp); \
struct tex_filter2D_ ## type ## _reader \
{ \
typedef type elem_type; \
typedef int index_type; \
const int xoff; \
const int yoff; \
tex_filter2D_ ## type ## _reader (int xoff_, int yoff_) : xoff(xoff_), yoff(yoff_) {} \
__device__ __forceinline__ elem_type operator ()(index_type y, index_type x) const \
{ \
return tex2D(tex_filter2D_ ## type , x + xoff, y + yoff); \
} \
}; \
template <typename D, template <typename> class Brd> struct Filter2DCaller< type , D, Brd> \
{ \
static void call(const PtrStepSz< type > srcWhole, int xoff, int yoff, PtrStepSz<D> dst, \
int kWidth, int kHeight, int anchorX, int anchorY, const float* borderValue, cudaStream_t stream) \
{ \
typedef typename TypeVec<float, VecTraits< type >::cn>::vec_type work_type; \
dim3 block(16, 16); \
dim3 grid(divUp(dst.cols, block.x), divUp(dst.rows, block.y)); \
bindTexture(&tex_filter2D_ ## type , srcWhole); \
tex_filter2D_ ## type ##_reader texSrc(xoff, yoff); \
Brd<work_type> brd(dst.rows, dst.cols, VecTraits<work_type>::make(borderValue)); \
BorderReader< tex_filter2D_ ## type ##_reader, Brd<work_type> > brdSrc(texSrc, brd); \
filter2D<<<grid, block, 0, stream>>>(brdSrc, dst, kWidth, kHeight, anchorX, anchorY); \
cudaSafeCall( cudaGetLastError() ); \
if (stream == 0) \
cudaSafeCall( cudaDeviceSynchronize() ); \
} \
};
IMPLEMENT_FILTER2D_TEX_READER(uchar);
IMPLEMENT_FILTER2D_TEX_READER(uchar4);
IMPLEMENT_FILTER2D_TEX_READER(ushort);
IMPLEMENT_FILTER2D_TEX_READER(ushort4);
IMPLEMENT_FILTER2D_TEX_READER(float);
IMPLEMENT_FILTER2D_TEX_READER(float4);
#undef IMPLEMENT_FILTER2D_TEX_READER
template <typename T, typename D>
void filter2D_gpu(PtrStepSzb srcWhole, int ofsX, int ofsY, PtrStepSzb dst,
int kWidth, int kHeight, int anchorX, int anchorY, const float* kernel,
int borderMode, const float* borderValue, cudaStream_t stream)
{
typedef void (*func_t)(const PtrStepSz<T> srcWhole, int xoff, int yoff, PtrStepSz<D> dst, int kWidth, int kHeight, int anchorX, int anchorY, const float* borderValue, cudaStream_t stream);
static const func_t funcs[] =
{
Filter2DCaller<T, D, BrdReflect101>::call,
Filter2DCaller<T, D, BrdReplicate>::call,
Filter2DCaller<T, D, BrdConstant>::call,
Filter2DCaller<T, D, BrdReflect>::call,
Filter2DCaller<T, D, BrdWrap>::call
};
if (stream == 0)
cudaSafeCall( cudaMemcpyToSymbol(c_filter2DKernel, kernel, kWidth * kHeight * sizeof(float), 0, cudaMemcpyDeviceToDevice) );
else
cudaSafeCall( cudaMemcpyToSymbolAsync(c_filter2DKernel, kernel, kWidth * kHeight * sizeof(float), 0, cudaMemcpyDeviceToDevice, stream) );
funcs[borderMode](static_cast< PtrStepSz<T> >(srcWhole), ofsX, ofsY, static_cast< PtrStepSz<D> >(dst), kWidth, kHeight, anchorX, anchorY, borderValue, stream);
}
template void filter2D_gpu<uchar, uchar>(PtrStepSzb srcWhole, int ofsX, int ofsY, PtrStepSzb dst, int kWidth, int kHeight, int anchorX, int anchorY, const float* kernel, int borderMode, const float* borderValue, cudaStream_t stream);
template void filter2D_gpu<uchar4, uchar4>(PtrStepSzb srcWhole, int ofsX, int ofsY, PtrStepSzb dst, int kWidth, int kHeight, int anchorX, int anchorY, const float* kernel, int borderMode, const float* borderValue, cudaStream_t stream);
template void filter2D_gpu<ushort, ushort>(PtrStepSzb srcWhole, int ofsX, int ofsY, PtrStepSzb dst, int kWidth, int kHeight, int anchorX, int anchorY, const float* kernel, int borderMode, const float* borderValue, cudaStream_t stream);
template void filter2D_gpu<ushort4, ushort4>(PtrStepSzb srcWhole, int ofsX, int ofsY, PtrStepSzb dst, int kWidth, int kHeight, int anchorX, int anchorY, const float* kernel, int borderMode, const float* borderValue, cudaStream_t stream);
template void filter2D_gpu<float, float>(PtrStepSzb srcWhole, int ofsX, int ofsY, PtrStepSzb dst, int kWidth, int kHeight, int anchorX, int anchorY, const float* kernel, int borderMode, const float* borderValue, cudaStream_t stream);
template void filter2D_gpu<float4, float4>(PtrStepSzb srcWhole, int ofsX, int ofsY, PtrStepSzb dst, int kWidth, int kHeight, int anchorX, int anchorY, const float* kernel, int borderMode, const float* borderValue, cudaStream_t stream);
} // namespace imgproc
}}} // namespace cv { namespace gpu { namespace cudev {

@ -0,0 +1,9 @@
if(ANDROID OR IOS)
ocv_module_disable(gpufilters)
endif()
set(the_description "GPU-accelerated Image Filtering")
ocv_warnings_disable(CMAKE_CXX_FLAGS /wd4127 /wd4324 /wd4512 -Wundef -Wmissing-declarations)
ocv_define_module(gpufilters opencv_imgproc OPTIONAL opencv_gpuarithm)

@ -0,0 +1,8 @@
*******************************************
gpufilters. GPU-accelerated Image Filtering
*******************************************
.. toctree::
:maxdepth: 1
filtering

@ -0,0 +1,269 @@
/*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*/
#ifndef __OPENCV_GPUFILTERS_HPP__
#define __OPENCV_GPUFILTERS_HPP__
#ifndef __cplusplus
# error gpufilters.hpp header must be compiled as C++
#endif
#include "opencv2/core/gpumat.hpp"
#include "opencv2/core/base.hpp"
namespace cv { namespace gpu {
/*!
The Base Class for 1D or Row-wise Filters
This is the base class for linear or non-linear filters that process 1D data.
In particular, such filters are used for the "horizontal" filtering parts in separable filters.
*/
class CV_EXPORTS BaseRowFilter_GPU
{
public:
BaseRowFilter_GPU(int ksize_, int anchor_) : ksize(ksize_), anchor(anchor_) {}
virtual ~BaseRowFilter_GPU() {}
virtual void operator()(const GpuMat& src, GpuMat& dst, Stream& stream = Stream::Null()) = 0;
int ksize, anchor;
};
/*!
The Base Class for Column-wise Filters
This is the base class for linear or non-linear filters that process columns of 2D arrays.
Such filters are used for the "vertical" filtering parts in separable filters.
*/
class CV_EXPORTS BaseColumnFilter_GPU
{
public:
BaseColumnFilter_GPU(int ksize_, int anchor_) : ksize(ksize_), anchor(anchor_) {}
virtual ~BaseColumnFilter_GPU() {}
virtual void operator()(const GpuMat& src, GpuMat& dst, Stream& stream = Stream::Null()) = 0;
int ksize, anchor;
};
/*!
The Base Class for Non-Separable 2D Filters.
This is the base class for linear or non-linear 2D filters.
*/
class CV_EXPORTS BaseFilter_GPU
{
public:
BaseFilter_GPU(const Size& ksize_, const Point& anchor_) : ksize(ksize_), anchor(anchor_) {}
virtual ~BaseFilter_GPU() {}
virtual void operator()(const GpuMat& src, GpuMat& dst, Stream& stream = Stream::Null()) = 0;
Size ksize;
Point anchor;
};
/*!
The Base Class for Filter Engine.
The class can be used to apply an arbitrary filtering operation to an image.
It contains all the necessary intermediate buffers.
*/
class CV_EXPORTS FilterEngine_GPU
{
public:
virtual ~FilterEngine_GPU() {}
virtual void apply(const GpuMat& src, GpuMat& dst, Rect roi = Rect(0,0,-1,-1), Stream& stream = Stream::Null()) = 0;
};
//! returns the non-separable filter engine with the specified filter
CV_EXPORTS Ptr<FilterEngine_GPU> createFilter2D_GPU(const Ptr<BaseFilter_GPU>& filter2D, int srcType, int dstType);
//! returns the separable filter engine with the specified filters
CV_EXPORTS Ptr<FilterEngine_GPU> createSeparableFilter_GPU(const Ptr<BaseRowFilter_GPU>& rowFilter,
const Ptr<BaseColumnFilter_GPU>& columnFilter, int srcType, int bufType, int dstType);
CV_EXPORTS Ptr<FilterEngine_GPU> createSeparableFilter_GPU(const Ptr<BaseRowFilter_GPU>& rowFilter,
const Ptr<BaseColumnFilter_GPU>& columnFilter, int srcType, int bufType, int dstType, GpuMat& buf);
//! returns horizontal 1D box filter
//! supports only CV_8UC1 source type and CV_32FC1 sum type
CV_EXPORTS Ptr<BaseRowFilter_GPU> getRowSumFilter_GPU(int srcType, int sumType, int ksize, int anchor = -1);
//! returns vertical 1D box filter
//! supports only CV_8UC1 sum type and CV_32FC1 dst type
CV_EXPORTS Ptr<BaseColumnFilter_GPU> getColumnSumFilter_GPU(int sumType, int dstType, int ksize, int anchor = -1);
//! returns 2D box filter
//! supports CV_8UC1 and CV_8UC4 source type, dst type must be the same as source type
CV_EXPORTS Ptr<BaseFilter_GPU> getBoxFilter_GPU(int srcType, int dstType, const Size& ksize, Point anchor = Point(-1, -1));
//! returns box filter engine
CV_EXPORTS Ptr<FilterEngine_GPU> createBoxFilter_GPU(int srcType, int dstType, const Size& ksize,
const Point& anchor = Point(-1,-1));
//! returns 2D morphological filter
//! only MORPH_ERODE and MORPH_DILATE are supported
//! supports CV_8UC1 and CV_8UC4 types
//! kernel must have CV_8UC1 type, one rows and cols == ksize.width * ksize.height
CV_EXPORTS Ptr<BaseFilter_GPU> getMorphologyFilter_GPU(int op, int type, const Mat& kernel, const Size& ksize,
Point anchor=Point(-1,-1));
//! returns morphological filter engine. Only MORPH_ERODE and MORPH_DILATE are supported.
CV_EXPORTS Ptr<FilterEngine_GPU> createMorphologyFilter_GPU(int op, int type, const Mat& kernel,
const Point& anchor = Point(-1,-1), int iterations = 1);
CV_EXPORTS Ptr<FilterEngine_GPU> createMorphologyFilter_GPU(int op, int type, const Mat& kernel, GpuMat& buf,
const Point& anchor = Point(-1,-1), int iterations = 1);
//! returns 2D filter with the specified kernel
//! supports CV_8U, CV_16U and CV_32F one and four channel image
CV_EXPORTS Ptr<BaseFilter_GPU> getLinearFilter_GPU(int srcType, int dstType, const Mat& kernel, Point anchor = Point(-1, -1), int borderType = BORDER_DEFAULT);
//! returns the non-separable linear filter engine
CV_EXPORTS Ptr<FilterEngine_GPU> createLinearFilter_GPU(int srcType, int dstType, const Mat& kernel,
Point anchor = Point(-1,-1), int borderType = BORDER_DEFAULT);
//! returns the primitive row filter with the specified kernel.
//! supports only CV_8UC1, CV_8UC4, CV_16SC1, CV_16SC2, CV_32SC1, CV_32FC1 source type.
//! there are two version of algorithm: NPP and OpenCV.
//! NPP calls when srcType == CV_8UC1 or srcType == CV_8UC4 and bufType == srcType,
//! otherwise calls OpenCV version.
//! NPP supports only BORDER_CONSTANT border type.
//! OpenCV version supports only CV_32F as buffer depth and
//! BORDER_REFLECT101, BORDER_REPLICATE and BORDER_CONSTANT border types.
CV_EXPORTS Ptr<BaseRowFilter_GPU> getLinearRowFilter_GPU(int srcType, int bufType, const Mat& rowKernel,
int anchor = -1, int borderType = BORDER_DEFAULT);
//! returns the primitive column filter with the specified kernel.
//! supports only CV_8UC1, CV_8UC4, CV_16SC1, CV_16SC2, CV_32SC1, CV_32FC1 dst type.
//! there are two version of algorithm: NPP and OpenCV.
//! NPP calls when dstType == CV_8UC1 or dstType == CV_8UC4 and bufType == dstType,
//! otherwise calls OpenCV version.
//! NPP supports only BORDER_CONSTANT border type.
//! OpenCV version supports only CV_32F as buffer depth and
//! BORDER_REFLECT101, BORDER_REPLICATE and BORDER_CONSTANT border types.
CV_EXPORTS Ptr<BaseColumnFilter_GPU> getLinearColumnFilter_GPU(int bufType, int dstType, const Mat& columnKernel,
int anchor = -1, int borderType = BORDER_DEFAULT);
//! returns the separable linear filter engine
CV_EXPORTS Ptr<FilterEngine_GPU> createSeparableLinearFilter_GPU(int srcType, int dstType, const Mat& rowKernel,
const Mat& columnKernel, const Point& anchor = Point(-1,-1), int rowBorderType = BORDER_DEFAULT,
int columnBorderType = -1);
CV_EXPORTS Ptr<FilterEngine_GPU> createSeparableLinearFilter_GPU(int srcType, int dstType, const Mat& rowKernel,
const Mat& columnKernel, GpuMat& buf, const Point& anchor = Point(-1,-1), int rowBorderType = BORDER_DEFAULT,
int columnBorderType = -1);
//! returns filter engine for the generalized Sobel operator
CV_EXPORTS Ptr<FilterEngine_GPU> createDerivFilter_GPU(int srcType, int dstType, int dx, int dy, int ksize,
int rowBorderType = BORDER_DEFAULT, int columnBorderType = -1);
CV_EXPORTS Ptr<FilterEngine_GPU> createDerivFilter_GPU(int srcType, int dstType, int dx, int dy, int ksize, GpuMat& buf,
int rowBorderType = BORDER_DEFAULT, int columnBorderType = -1);
//! returns the Gaussian filter engine
CV_EXPORTS Ptr<FilterEngine_GPU> createGaussianFilter_GPU(int type, Size ksize, double sigma1, double sigma2 = 0,
int rowBorderType = BORDER_DEFAULT, int columnBorderType = -1);
CV_EXPORTS Ptr<FilterEngine_GPU> createGaussianFilter_GPU(int type, Size ksize, GpuMat& buf, double sigma1, double sigma2 = 0,
int rowBorderType = BORDER_DEFAULT, int columnBorderType = -1);
//! returns maximum filter
CV_EXPORTS Ptr<BaseFilter_GPU> getMaxFilter_GPU(int srcType, int dstType, const Size& ksize, Point anchor = Point(-1,-1));
//! returns minimum filter
CV_EXPORTS Ptr<BaseFilter_GPU> getMinFilter_GPU(int srcType, int dstType, const Size& ksize, Point anchor = Point(-1,-1));
//! smooths the image using the normalized box filter
//! supports CV_8UC1, CV_8UC4 types
CV_EXPORTS void boxFilter(const GpuMat& src, GpuMat& dst, int ddepth, Size ksize, Point anchor = Point(-1,-1), Stream& stream = Stream::Null());
//! a synonym for normalized box filter
static inline void blur(const GpuMat& src, GpuMat& dst, Size ksize, Point anchor = Point(-1,-1), Stream& stream = Stream::Null())
{
boxFilter(src, dst, -1, ksize, anchor, stream);
}
//! erodes the image (applies the local minimum operator)
CV_EXPORTS void erode(const GpuMat& src, GpuMat& dst, const Mat& kernel, Point anchor = Point(-1, -1), int iterations = 1);
CV_EXPORTS void erode(const GpuMat& src, GpuMat& dst, const Mat& kernel, GpuMat& buf,
Point anchor = Point(-1, -1), int iterations = 1,
Stream& stream = Stream::Null());
//! dilates the image (applies the local maximum operator)
CV_EXPORTS void dilate(const GpuMat& src, GpuMat& dst, const Mat& kernel, Point anchor = Point(-1, -1), int iterations = 1);
CV_EXPORTS void dilate(const GpuMat& src, GpuMat& dst, const Mat& kernel, GpuMat& buf,
Point anchor = Point(-1, -1), int iterations = 1,
Stream& stream = Stream::Null());
//! applies an advanced morphological operation to the image
CV_EXPORTS void morphologyEx(const GpuMat& src, GpuMat& dst, int op, const Mat& kernel, Point anchor = Point(-1, -1), int iterations = 1);
CV_EXPORTS void morphologyEx(const GpuMat& src, GpuMat& dst, int op, const Mat& kernel, GpuMat& buf1, GpuMat& buf2,
Point anchor = Point(-1, -1), int iterations = 1, Stream& stream = Stream::Null());
//! applies non-separable 2D linear filter to the image
CV_EXPORTS void filter2D(const GpuMat& src, GpuMat& dst, int ddepth, const Mat& kernel, Point anchor=Point(-1,-1), int borderType = BORDER_DEFAULT, Stream& stream = Stream::Null());
//! applies separable 2D linear filter to the image
CV_EXPORTS void sepFilter2D(const GpuMat& src, GpuMat& dst, int ddepth, const Mat& kernelX, const Mat& kernelY,
Point anchor = Point(-1,-1), int rowBorderType = BORDER_DEFAULT, int columnBorderType = -1);
CV_EXPORTS void sepFilter2D(const GpuMat& src, GpuMat& dst, int ddepth, const Mat& kernelX, const Mat& kernelY, GpuMat& buf,
Point anchor = Point(-1,-1), int rowBorderType = BORDER_DEFAULT, int columnBorderType = -1,
Stream& stream = Stream::Null());
//! applies generalized Sobel operator to the image
CV_EXPORTS void Sobel(const GpuMat& src, GpuMat& dst, int ddepth, int dx, int dy, int ksize = 3, double scale = 1,
int rowBorderType = BORDER_DEFAULT, int columnBorderType = -1);
CV_EXPORTS void Sobel(const GpuMat& src, GpuMat& dst, int ddepth, int dx, int dy, GpuMat& buf, int ksize = 3, double scale = 1,
int rowBorderType = BORDER_DEFAULT, int columnBorderType = -1, Stream& stream = Stream::Null());
//! applies the vertical or horizontal Scharr operator to the image
CV_EXPORTS void Scharr(const GpuMat& src, GpuMat& dst, int ddepth, int dx, int dy, double scale = 1,
int rowBorderType = BORDER_DEFAULT, int columnBorderType = -1);
CV_EXPORTS void Scharr(const GpuMat& src, GpuMat& dst, int ddepth, int dx, int dy, GpuMat& buf, double scale = 1,
int rowBorderType = BORDER_DEFAULT, int columnBorderType = -1, Stream& stream = Stream::Null());
//! smooths the image using Gaussian filter.
CV_EXPORTS void GaussianBlur(const GpuMat& src, GpuMat& dst, Size ksize, double sigma1, double sigma2 = 0,
int rowBorderType = BORDER_DEFAULT, int columnBorderType = -1);
CV_EXPORTS void GaussianBlur(const GpuMat& src, GpuMat& dst, Size ksize, GpuMat& buf, double sigma1, double sigma2 = 0,
int rowBorderType = BORDER_DEFAULT, int columnBorderType = -1, Stream& stream = Stream::Null());
//! applies Laplacian operator to the image
//! supports only ksize = 1 and ksize = 3
CV_EXPORTS void Laplacian(const GpuMat& src, GpuMat& dst, int ddepth, int ksize = 1, double scale = 1, int borderType = BORDER_DEFAULT, Stream& stream = Stream::Null());
}} // namespace cv { namespace gpu {
#endif /* __OPENCV_GPUFILTERS_HPP__ */

@ -51,7 +51,7 @@ using namespace perf;
DEF_PARAM_TEST(Sz_Type_KernelSz, cv::Size, MatType, int);
PERF_TEST_P(Sz_Type_KernelSz, Filters_Blur,
PERF_TEST_P(Sz_Type_KernelSz, Blur,
Combine(GPU_TYPICAL_MAT_SIZES,
Values(CV_8UC1, CV_8UC4),
Values(3, 5, 7)))
@ -87,7 +87,7 @@ PERF_TEST_P(Sz_Type_KernelSz, Filters_Blur,
//////////////////////////////////////////////////////////////////////
// Sobel
PERF_TEST_P(Sz_Type_KernelSz, Filters_Sobel, Combine(GPU_TYPICAL_MAT_SIZES, Values(CV_8UC1, CV_8UC4, CV_32FC1), Values(3, 5, 7, 9, 11, 13, 15)))
PERF_TEST_P(Sz_Type_KernelSz, Sobel, Combine(GPU_TYPICAL_MAT_SIZES, Values(CV_8UC1, CV_8UC4, CV_32FC1), Values(3, 5, 7, 9, 11, 13, 15)))
{
declare.time(20.0);
@ -121,7 +121,7 @@ PERF_TEST_P(Sz_Type_KernelSz, Filters_Sobel, Combine(GPU_TYPICAL_MAT_SIZES, Valu
//////////////////////////////////////////////////////////////////////
// Scharr
PERF_TEST_P(Sz_Type, Filters_Scharr, Combine(GPU_TYPICAL_MAT_SIZES, Values(CV_8UC1, CV_8UC4, CV_32FC1)))
PERF_TEST_P(Sz_Type, Scharr, Combine(GPU_TYPICAL_MAT_SIZES, Values(CV_8UC1, CV_8UC4, CV_32FC1)))
{
declare.time(20.0);
@ -154,7 +154,7 @@ PERF_TEST_P(Sz_Type, Filters_Scharr, Combine(GPU_TYPICAL_MAT_SIZES, Values(CV_8U
//////////////////////////////////////////////////////////////////////
// GaussianBlur
PERF_TEST_P(Sz_Type_KernelSz, Filters_GaussianBlur, Combine(GPU_TYPICAL_MAT_SIZES, Values(CV_8UC1, CV_8UC4, CV_32FC1), Values(3, 5, 7, 9, 11, 13, 15)))
PERF_TEST_P(Sz_Type_KernelSz, GaussianBlur, Combine(GPU_TYPICAL_MAT_SIZES, Values(CV_8UC1, CV_8UC4, CV_32FC1), Values(3, 5, 7, 9, 11, 13, 15)))
{
declare.time(20.0);
@ -188,7 +188,7 @@ PERF_TEST_P(Sz_Type_KernelSz, Filters_GaussianBlur, Combine(GPU_TYPICAL_MAT_SIZE
//////////////////////////////////////////////////////////////////////
// Laplacian
PERF_TEST_P(Sz_Type_KernelSz, Filters_Laplacian, Combine(GPU_TYPICAL_MAT_SIZES, Values(CV_8UC1, CV_8UC4, CV_32FC1, CV_32FC4), Values(1, 3)))
PERF_TEST_P(Sz_Type_KernelSz, Laplacian, Combine(GPU_TYPICAL_MAT_SIZES, Values(CV_8UC1, CV_8UC4, CV_32FC1, CV_32FC4), Values(1, 3)))
{
declare.time(20.0);
@ -221,7 +221,7 @@ PERF_TEST_P(Sz_Type_KernelSz, Filters_Laplacian, Combine(GPU_TYPICAL_MAT_SIZES,
//////////////////////////////////////////////////////////////////////
// Erode
PERF_TEST_P(Sz_Type, Filters_Erode, Combine(GPU_TYPICAL_MAT_SIZES, Values(CV_8UC1, CV_8UC4)))
PERF_TEST_P(Sz_Type, Erode, Combine(GPU_TYPICAL_MAT_SIZES, Values(CV_8UC1, CV_8UC4)))
{
declare.time(20.0);
@ -256,7 +256,7 @@ PERF_TEST_P(Sz_Type, Filters_Erode, Combine(GPU_TYPICAL_MAT_SIZES, Values(CV_8UC
//////////////////////////////////////////////////////////////////////
// Dilate
PERF_TEST_P(Sz_Type, Filters_Dilate, Combine(GPU_TYPICAL_MAT_SIZES, Values(CV_8UC1, CV_8UC4)))
PERF_TEST_P(Sz_Type, Dilate, Combine(GPU_TYPICAL_MAT_SIZES, Values(CV_8UC1, CV_8UC4)))
{
declare.time(20.0);
@ -295,7 +295,7 @@ CV_ENUM(MorphOp, MORPH_OPEN, MORPH_CLOSE, MORPH_GRADIENT, MORPH_TOPHAT, MORPH_BL
DEF_PARAM_TEST(Sz_Type_Op, cv::Size, MatType, MorphOp);
PERF_TEST_P(Sz_Type_Op, Filters_MorphologyEx, Combine(GPU_TYPICAL_MAT_SIZES, Values(CV_8UC1, CV_8UC4), MorphOp::all()))
PERF_TEST_P(Sz_Type_Op, MorphologyEx, Combine(GPU_TYPICAL_MAT_SIZES, Values(CV_8UC1, CV_8UC4), MorphOp::all()))
{
declare.time(20.0);
@ -332,7 +332,7 @@ PERF_TEST_P(Sz_Type_Op, Filters_MorphologyEx, Combine(GPU_TYPICAL_MAT_SIZES, Val
//////////////////////////////////////////////////////////////////////
// Filter2D
PERF_TEST_P(Sz_Type_KernelSz, Filters_Filter2D, Combine(GPU_TYPICAL_MAT_SIZES, Values(CV_8UC1, CV_8UC4, CV_32FC1, CV_32FC4), Values(3, 5, 7, 9, 11, 13, 15)))
PERF_TEST_P(Sz_Type_KernelSz, Filter2D, Combine(GPU_TYPICAL_MAT_SIZES, Values(CV_8UC1, CV_8UC4, CV_32FC1, CV_32FC4), Values(3, 5, 7, 9, 11, 13, 15)))
{
declare.time(20.0);

@ -0,0 +1,47 @@
/*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 "perf_precomp.hpp"
using namespace perf;
CV_PERF_TEST_MAIN(gpufilters, printCudaInfo())

@ -0,0 +1,43 @@
/*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 "perf_precomp.hpp"

@ -0,0 +1,64 @@
/*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*/
#ifdef __GNUC__
# pragma GCC diagnostic ignored "-Wmissing-declarations"
# if defined __clang__ || defined __APPLE__
# pragma GCC diagnostic ignored "-Wmissing-prototypes"
# pragma GCC diagnostic ignored "-Wextra"
# endif
#endif
#ifndef __OPENCV_PERF_PRECOMP_HPP__
#define __OPENCV_PERF_PRECOMP_HPP__
#include "opencv2/ts.hpp"
#include "opencv2/ts/gpu_perf.hpp"
#include "opencv2/gpufilters.hpp"
#include "opencv2/imgproc.hpp"
#ifdef GTEST_CREATE_SHARED_LIBRARY
#error no modules except ts should have GTEST_CREATE_SHARED_LIBRARY defined
#endif
#endif

@ -42,7 +42,7 @@
#if !defined CUDA_DISABLER
#include "column_filter.h"
#include "column_filter.hpp"
namespace filter
{

@ -42,7 +42,7 @@
#if !defined CUDA_DISABLER
#include "column_filter.h"
#include "column_filter.hpp"
namespace filter
{

@ -42,7 +42,7 @@
#if !defined CUDA_DISABLER
#include "column_filter.h"
#include "column_filter.hpp"
namespace filter
{

@ -42,7 +42,7 @@
#if !defined CUDA_DISABLER
#include "column_filter.h"
#include "column_filter.hpp"
namespace filter
{

@ -42,7 +42,7 @@
#if !defined CUDA_DISABLER
#include "column_filter.h"
#include "column_filter.hpp"
namespace filter
{

@ -42,7 +42,7 @@
#if !defined CUDA_DISABLER
#include "column_filter.h"
#include "column_filter.hpp"
namespace filter
{

@ -42,7 +42,7 @@
#if !defined CUDA_DISABLER
#include "column_filter.h"
#include "column_filter.hpp"
namespace filter
{

@ -42,7 +42,7 @@
#if !defined CUDA_DISABLER
#include "column_filter.h"
#include "column_filter.hpp"
namespace filter
{

@ -42,7 +42,7 @@
#if !defined CUDA_DISABLER
#include "column_filter.h"
#include "column_filter.hpp"
namespace filter
{

@ -42,7 +42,7 @@
#if !defined CUDA_DISABLER
#include "column_filter.h"
#include "column_filter.hpp"
namespace filter
{

@ -42,7 +42,7 @@
#if !defined CUDA_DISABLER
#include "column_filter.h"
#include "column_filter.hpp"
namespace filter
{

@ -42,7 +42,7 @@
#if !defined CUDA_DISABLER
#include "column_filter.h"
#include "column_filter.hpp"
namespace filter
{

@ -42,7 +42,7 @@
#if !defined CUDA_DISABLER
#include "column_filter.h"
#include "column_filter.hpp"
namespace filter
{

@ -42,7 +42,7 @@
#if !defined CUDA_DISABLER
#include "column_filter.h"
#include "column_filter.hpp"
namespace filter
{

@ -42,7 +42,7 @@
#if !defined CUDA_DISABLER
#include "column_filter.h"
#include "column_filter.hpp"
namespace filter
{

@ -0,0 +1,158 @@
/*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*/
#if !defined CUDA_DISABLER
#include "opencv2/core/cuda/common.hpp"
#include "opencv2/core/cuda/saturate_cast.hpp"
#include "opencv2/core/cuda/border_interpolate.hpp"
namespace cv { namespace gpu { namespace cudev
{
namespace imgproc
{
#define FILTER2D_MAX_KERNEL_SIZE 16
__constant__ float c_filter2DKernel[FILTER2D_MAX_KERNEL_SIZE * FILTER2D_MAX_KERNEL_SIZE];
template <class SrcT, typename D>
__global__ void filter2D(const SrcT src, PtrStepSz<D> dst, const int kWidth, const int kHeight, const int anchorX, const int anchorY)
{
typedef typename TypeVec<float, VecTraits<D>::cn>::vec_type sum_t;
const int x = blockIdx.x * blockDim.x + threadIdx.x;
const int y = blockIdx.y * blockDim.y + threadIdx.y;
if (x >= dst.cols || y >= dst.rows)
return;
sum_t res = VecTraits<sum_t>::all(0);
int kInd = 0;
for (int i = 0; i < kHeight; ++i)
{
for (int j = 0; j < kWidth; ++j)
res = res + src(y - anchorY + i, x - anchorX + j) * c_filter2DKernel[kInd++];
}
dst(y, x) = saturate_cast<D>(res);
}
template <typename T, typename D, template <typename> class Brd> struct Filter2DCaller;
#define IMPLEMENT_FILTER2D_TEX_READER(type) \
texture< type , cudaTextureType2D, cudaReadModeElementType> tex_filter2D_ ## type (0, cudaFilterModePoint, cudaAddressModeClamp); \
struct tex_filter2D_ ## type ## _reader \
{ \
typedef type elem_type; \
typedef int index_type; \
const int xoff; \
const int yoff; \
tex_filter2D_ ## type ## _reader (int xoff_, int yoff_) : xoff(xoff_), yoff(yoff_) {} \
__device__ __forceinline__ elem_type operator ()(index_type y, index_type x) const \
{ \
return tex2D(tex_filter2D_ ## type , x + xoff, y + yoff); \
} \
}; \
template <typename D, template <typename> class Brd> struct Filter2DCaller< type , D, Brd> \
{ \
static void call(const PtrStepSz< type > srcWhole, int xoff, int yoff, PtrStepSz<D> dst, \
int kWidth, int kHeight, int anchorX, int anchorY, const float* borderValue, cudaStream_t stream) \
{ \
typedef typename TypeVec<float, VecTraits< type >::cn>::vec_type work_type; \
dim3 block(16, 16); \
dim3 grid(divUp(dst.cols, block.x), divUp(dst.rows, block.y)); \
bindTexture(&tex_filter2D_ ## type , srcWhole); \
tex_filter2D_ ## type ##_reader texSrc(xoff, yoff); \
Brd<work_type> brd(dst.rows, dst.cols, VecTraits<work_type>::make(borderValue)); \
BorderReader< tex_filter2D_ ## type ##_reader, Brd<work_type> > brdSrc(texSrc, brd); \
filter2D<<<grid, block, 0, stream>>>(brdSrc, dst, kWidth, kHeight, anchorX, anchorY); \
cudaSafeCall( cudaGetLastError() ); \
if (stream == 0) \
cudaSafeCall( cudaDeviceSynchronize() ); \
} \
};
IMPLEMENT_FILTER2D_TEX_READER(uchar);
IMPLEMENT_FILTER2D_TEX_READER(uchar4);
IMPLEMENT_FILTER2D_TEX_READER(ushort);
IMPLEMENT_FILTER2D_TEX_READER(ushort4);
IMPLEMENT_FILTER2D_TEX_READER(float);
IMPLEMENT_FILTER2D_TEX_READER(float4);
#undef IMPLEMENT_FILTER2D_TEX_READER
template <typename T, typename D>
void filter2D_gpu(PtrStepSzb srcWhole, int ofsX, int ofsY, PtrStepSzb dst,
int kWidth, int kHeight, int anchorX, int anchorY, const float* kernel,
int borderMode, const float* borderValue, cudaStream_t stream)
{
typedef void (*func_t)(const PtrStepSz<T> srcWhole, int xoff, int yoff, PtrStepSz<D> dst, int kWidth, int kHeight, int anchorX, int anchorY, const float* borderValue, cudaStream_t stream);
static const func_t funcs[] =
{
Filter2DCaller<T, D, BrdReflect101>::call,
Filter2DCaller<T, D, BrdReplicate>::call,
Filter2DCaller<T, D, BrdConstant>::call,
Filter2DCaller<T, D, BrdReflect>::call,
Filter2DCaller<T, D, BrdWrap>::call
};
if (stream == 0)
cudaSafeCall( cudaMemcpyToSymbol(c_filter2DKernel, kernel, kWidth * kHeight * sizeof(float), 0, cudaMemcpyDeviceToDevice) );
else
cudaSafeCall( cudaMemcpyToSymbolAsync(c_filter2DKernel, kernel, kWidth * kHeight * sizeof(float), 0, cudaMemcpyDeviceToDevice, stream) );
funcs[borderMode](static_cast< PtrStepSz<T> >(srcWhole), ofsX, ofsY, static_cast< PtrStepSz<D> >(dst), kWidth, kHeight, anchorX, anchorY, borderValue, stream);
}
template void filter2D_gpu<uchar, uchar>(PtrStepSzb srcWhole, int ofsX, int ofsY, PtrStepSzb dst, int kWidth, int kHeight, int anchorX, int anchorY, const float* kernel, int borderMode, const float* borderValue, cudaStream_t stream);
template void filter2D_gpu<uchar4, uchar4>(PtrStepSzb srcWhole, int ofsX, int ofsY, PtrStepSzb dst, int kWidth, int kHeight, int anchorX, int anchorY, const float* kernel, int borderMode, const float* borderValue, cudaStream_t stream);
template void filter2D_gpu<ushort, ushort>(PtrStepSzb srcWhole, int ofsX, int ofsY, PtrStepSzb dst, int kWidth, int kHeight, int anchorX, int anchorY, const float* kernel, int borderMode, const float* borderValue, cudaStream_t stream);
template void filter2D_gpu<ushort4, ushort4>(PtrStepSzb srcWhole, int ofsX, int ofsY, PtrStepSzb dst, int kWidth, int kHeight, int anchorX, int anchorY, const float* kernel, int borderMode, const float* borderValue, cudaStream_t stream);
template void filter2D_gpu<float, float>(PtrStepSzb srcWhole, int ofsX, int ofsY, PtrStepSzb dst, int kWidth, int kHeight, int anchorX, int anchorY, const float* kernel, int borderMode, const float* borderValue, cudaStream_t stream);
template void filter2D_gpu<float4, float4>(PtrStepSzb srcWhole, int ofsX, int ofsY, PtrStepSzb dst, int kWidth, int kHeight, int anchorX, int anchorY, const float* kernel, int borderMode, const float* borderValue, cudaStream_t stream);
}
}}}
#endif // CUDA_DISABLER

@ -42,7 +42,7 @@
#if !defined CUDA_DISABLER
#include "row_filter.h"
#include "row_filter.hpp"
namespace filter
{

@ -42,7 +42,7 @@
#if !defined CUDA_DISABLER
#include "row_filter.h"
#include "row_filter.hpp"
namespace filter
{

@ -42,7 +42,7 @@
#if !defined CUDA_DISABLER
#include "row_filter.h"
#include "row_filter.hpp"
namespace filter
{

@ -42,7 +42,7 @@
#if !defined CUDA_DISABLER
#include "row_filter.h"
#include "row_filter.hpp"
namespace filter
{

@ -42,7 +42,7 @@
#if !defined CUDA_DISABLER
#include "row_filter.h"
#include "row_filter.hpp"
namespace filter
{

@ -42,7 +42,7 @@
#if !defined CUDA_DISABLER
#include "row_filter.h"
#include "row_filter.hpp"
namespace filter
{

@ -42,7 +42,7 @@
#if !defined CUDA_DISABLER
#include "row_filter.h"
#include "row_filter.hpp"
namespace filter
{

@ -42,7 +42,7 @@
#if !defined CUDA_DISABLER
#include "row_filter.h"
#include "row_filter.hpp"
namespace filter
{

@ -42,7 +42,7 @@
#if !defined CUDA_DISABLER
#include "row_filter.h"
#include "row_filter.hpp"
namespace filter
{

@ -42,7 +42,7 @@
#if !defined CUDA_DISABLER
#include "row_filter.h"
#include "row_filter.hpp"
namespace filter
{

@ -42,7 +42,7 @@
#if !defined CUDA_DISABLER
#include "row_filter.h"
#include "row_filter.hpp"
namespace filter
{

@ -42,7 +42,7 @@
#if !defined CUDA_DISABLER
#include "row_filter.h"
#include "row_filter.hpp"
namespace filter
{

@ -42,7 +42,7 @@
#if !defined CUDA_DISABLER
#include "row_filter.h"
#include "row_filter.hpp"
namespace filter
{

@ -42,7 +42,7 @@
#if !defined CUDA_DISABLER
#include "row_filter.h"
#include "row_filter.hpp"
namespace filter
{

@ -42,7 +42,7 @@
#if !defined CUDA_DISABLER
#include "row_filter.h"
#include "row_filter.hpp"
namespace filter
{

@ -45,7 +45,6 @@
using namespace cv;
using namespace cv::gpu;
#if !defined (HAVE_CUDA) || defined (CUDA_DISABLER)
Ptr<FilterEngine_GPU> cv::gpu::createFilter2D_GPU(const Ptr<BaseFilter_GPU>&, int, int) { throw_no_cuda(); return Ptr<FilterEngine_GPU>(0); }
@ -628,31 +627,44 @@ void cv::gpu::morphologyEx(const GpuMat& src, GpuMat& dst, int op, const Mat& ke
{
switch( op )
{
case MORPH_ERODE: erode(src, dst, kernel, buf1, anchor, iterations, stream); break;
case MORPH_DILATE: dilate(src, dst, kernel, buf1, anchor, iterations, stream); break;
case MORPH_ERODE:
erode(src, dst, kernel, buf1, anchor, iterations, stream);
break;
case MORPH_DILATE:
dilate(src, dst, kernel, buf1, anchor, iterations, stream);
break;
case MORPH_OPEN:
erode(src, buf2, kernel, buf1, anchor, iterations, stream);
dilate(buf2, dst, kernel, buf1, anchor, iterations, stream);
break;
case MORPH_CLOSE:
dilate(src, buf2, kernel, buf1, anchor, iterations, stream);
erode(buf2, dst, kernel, buf1, anchor, iterations, stream);
break;
#ifdef HAVE_OPENCV_GPUARITHM
case MORPH_GRADIENT:
erode(src, buf2, kernel, buf1, anchor, iterations, stream);
dilate(src, dst, kernel, buf1, anchor, iterations, stream);
subtract(dst, buf2, dst, GpuMat(), -1, stream);
gpu::subtract(dst, buf2, dst, GpuMat(), -1, stream);
break;
case MORPH_TOPHAT:
erode(src, dst, kernel, buf1, anchor, iterations, stream);
dilate(dst, buf2, kernel, buf1, anchor, iterations, stream);
subtract(src, buf2, dst, GpuMat(), -1, stream);
gpu::subtract(src, buf2, dst, GpuMat(), -1, stream);
break;
case MORPH_BLACKHAT:
dilate(src, dst, kernel, buf1, anchor, iterations, stream);
erode(dst, buf2, kernel, buf1, anchor, iterations, stream);
subtract(buf2, src, dst, GpuMat(), -1, stream);
gpu::subtract(buf2, src, dst, GpuMat(), -1, stream);
break;
#endif
default:
CV_Error(cv::Error::StsBadArg, "unknown morphological operation");
}

@ -0,0 +1,43 @@
/*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"

@ -0,0 +1,59 @@
/*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*/
#ifndef __OPENCV_PRECOMP_H__
#define __OPENCV_PRECOMP_H__
#include <limits>
#include "opencv2/gpufilters.hpp"
#include "opencv2/imgproc.hpp"
#include "opencv2/core/gpu_private.hpp"
#include "opencv2/opencv_modules.hpp"
#ifdef HAVE_OPENCV_GPUARITHM
# include "opencv2/gpuarithm.hpp"
#endif
#endif /* __OPENCV_PRECOMP_H__ */

@ -105,7 +105,7 @@ GPU_TEST_P(Blur, Accuracy)
EXPECT_MAT_NEAR(getInnerROI(dst_gold, ksize), getInnerROI(dst, ksize), 1.0);
}
INSTANTIATE_TEST_CASE_P(GPU_Filter, Blur, testing::Combine(
INSTANTIATE_TEST_CASE_P(GPU_Filters, Blur, testing::Combine(
ALL_DEVICES,
DIFFERENT_SIZES,
testing::Values(MatType(CV_8UC1), MatType(CV_8UC4)),
@ -164,7 +164,7 @@ GPU_TEST_P(Sobel, Accuracy)
EXPECT_MAT_NEAR(getInnerROI(dst_gold, ksize), getInnerROI(dst, ksize), CV_MAT_DEPTH(type) < CV_32F ? 0.0 : 0.1);
}
INSTANTIATE_TEST_CASE_P(GPU_Filter, Sobel, testing::Combine(
INSTANTIATE_TEST_CASE_P(GPU_Filters, Sobel, testing::Combine(
ALL_DEVICES,
DIFFERENT_SIZES,
testing::Values(MatDepth(CV_8U), MatDepth(CV_16U), MatDepth(CV_16S), MatDepth(CV_32F)),
@ -227,7 +227,7 @@ GPU_TEST_P(Scharr, Accuracy)
EXPECT_MAT_NEAR(getInnerROI(dst_gold, cv::Size(3, 3)), getInnerROI(dst, cv::Size(3, 3)), CV_MAT_DEPTH(type) < CV_32F ? 0.0 : 0.1);
}
INSTANTIATE_TEST_CASE_P(GPU_Filter, Scharr, testing::Combine(
INSTANTIATE_TEST_CASE_P(GPU_Filters, Scharr, testing::Combine(
ALL_DEVICES,
DIFFERENT_SIZES,
testing::Values(MatDepth(CV_8U), MatDepth(CV_16U), MatDepth(CV_16S), MatDepth(CV_32F)),
@ -301,7 +301,7 @@ GPU_TEST_P(GaussianBlur, Accuracy)
}
}
INSTANTIATE_TEST_CASE_P(GPU_Filter, GaussianBlur, testing::Combine(
INSTANTIATE_TEST_CASE_P(GPU_Filters, GaussianBlur, testing::Combine(
ALL_DEVICES,
DIFFERENT_SIZES,
testing::Values(MatDepth(CV_8U), MatDepth(CV_16U), MatDepth(CV_16S), MatDepth(CV_32F)),
@ -363,7 +363,7 @@ GPU_TEST_P(Laplacian, Accuracy)
EXPECT_MAT_NEAR(dst_gold, dst, src.depth() < CV_32F ? 0.0 : 1e-3);
}
INSTANTIATE_TEST_CASE_P(GPU_Filter, Laplacian, testing::Combine(
INSTANTIATE_TEST_CASE_P(GPU_Filters, Laplacian, testing::Combine(
ALL_DEVICES,
DIFFERENT_SIZES,
testing::Values(MatType(CV_8UC1), MatType(CV_8UC4), MatType(CV_32FC1)),
@ -411,7 +411,7 @@ GPU_TEST_P(Erode, Accuracy)
EXPECT_MAT_NEAR(getInnerROI(dst_gold, ksize), getInnerROI(dst, ksize), 0.0);
}
INSTANTIATE_TEST_CASE_P(GPU_Filter, Erode, testing::Combine(
INSTANTIATE_TEST_CASE_P(GPU_Filters, Erode, testing::Combine(
ALL_DEVICES,
DIFFERENT_SIZES,
testing::Values(MatType(CV_8UC1), MatType(CV_8UC4)),
@ -460,7 +460,7 @@ GPU_TEST_P(Dilate, Accuracy)
EXPECT_MAT_NEAR(getInnerROI(dst_gold, ksize), getInnerROI(dst, ksize), 0.0);
}
INSTANTIATE_TEST_CASE_P(GPU_Filter, Dilate, testing::Combine(
INSTANTIATE_TEST_CASE_P(GPU_Filters, Dilate, testing::Combine(
ALL_DEVICES,
DIFFERENT_SIZES,
testing::Values(MatType(CV_8UC1), MatType(CV_8UC4)),
@ -513,7 +513,7 @@ GPU_TEST_P(MorphEx, Accuracy)
EXPECT_MAT_NEAR(getInnerROI(dst_gold, border), getInnerROI(dst, border), 0.0);
}
INSTANTIATE_TEST_CASE_P(GPU_Filter, MorphEx, testing::Combine(
INSTANTIATE_TEST_CASE_P(GPU_Filters, MorphEx, testing::Combine(
ALL_DEVICES,
DIFFERENT_SIZES,
testing::Values(MatType(CV_8UC1), MatType(CV_8UC4)),
@ -565,7 +565,7 @@ GPU_TEST_P(Filter2D, Accuracy)
EXPECT_MAT_NEAR(dst_gold, dst, CV_MAT_DEPTH(type) == CV_32F ? 1e-1 : 1.0);
}
INSTANTIATE_TEST_CASE_P(GPU_Filter, Filter2D, testing::Combine(
INSTANTIATE_TEST_CASE_P(GPU_Filters, Filter2D, testing::Combine(
ALL_DEVICES,
DIFFERENT_SIZES,
testing::Values(MatType(CV_8UC1), MatType(CV_8UC4), MatType(CV_16UC1), MatType(CV_16UC4), MatType(CV_32FC1), MatType(CV_32FC4)),

@ -0,0 +1,120 @@
/*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 "test_precomp.hpp"
#ifdef HAVE_CUDA
using namespace std;
using namespace cv;
using namespace cv::gpu;
using namespace cvtest;
using namespace testing;
int main(int argc, char** argv)
{
try
{
const std::string keys =
"{ h help ? | | Print help}"
"{ i info | | Print information about system and exit }"
"{ device | -1 | Device on which tests will be executed (-1 means all devices) }"
;
CommandLineParser cmd(argc, (const char**)argv, keys);
if (cmd.has("help"))
{
cmd.printMessage();
return 0;
}
printCudaInfo();
if (cmd.has("info"))
{
return 0;
}
int device = cmd.get<int>("device");
if (device < 0)
{
DeviceManager::instance().loadAll();
cout << "Run tests on all supported devices \n" << endl;
}
else
{
DeviceManager::instance().load(device);
DeviceInfo info(device);
cout << "Run tests on device " << device << " [" << info.name() << "] \n" << endl;
}
TS::ptr()->init("gpu");
InitGoogleTest(&argc, argv);
return RUN_ALL_TESTS();
}
catch (const exception& e)
{
cerr << e.what() << endl;
return -1;
}
catch (...)
{
cerr << "Unknown error" << endl;
return -1;
}
return 0;
}
#else // HAVE_CUDA
int main()
{
printf("OpenCV was built without CUDA support\n");
return 0;
}
#endif // HAVE_CUDA

@ -0,0 +1,43 @@
/*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 "test_precomp.hpp"

@ -0,0 +1,60 @@
/*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*/
#ifdef __GNUC__
# pragma GCC diagnostic ignored "-Wmissing-declarations"
# if defined __clang__ || defined __APPLE__
# pragma GCC diagnostic ignored "-Wmissing-prototypes"
# pragma GCC diagnostic ignored "-Wextra"
# endif
#endif
#ifndef __OPENCV_TEST_PRECOMP_HPP__
#define __OPENCV_TEST_PRECOMP_HPP__
#include "opencv2/ts.hpp"
#include "opencv2/ts/gpu_test.hpp"
#include "opencv2/gpufilters.hpp"
#include "opencv2/imgproc.hpp"
#endif

@ -1,3 +1,3 @@
set(the_description "Images stitching")
ocv_define_module(stitching opencv_imgproc opencv_features2d opencv_calib3d opencv_objdetect OPTIONAL opencv_gpu opencv_gpuarithm opencv_nonfree)
ocv_define_module(stitching opencv_imgproc opencv_features2d opencv_calib3d opencv_objdetect OPTIONAL opencv_gpu opencv_gpuarithm opencv_gpufilters opencv_nonfree)

@ -4,4 +4,4 @@ endif()
set(the_description "Super Resolution")
ocv_warnings_disable(CMAKE_CXX_FLAGS /wd4127 -Wundef)
ocv_define_module(superres opencv_imgproc opencv_video OPTIONAL opencv_highgui opencv_gpu opencv_gpucodec opencv_gpuarithm)
ocv_define_module(superres opencv_imgproc opencv_video OPTIONAL opencv_highgui opencv_gpu opencv_gpuarithm opencv_gpufilters opencv_gpucodec)

@ -18,6 +18,7 @@ if(BUILD_EXAMPLES AND OCV_DEPENDENCIES_FOUND)
if(HAVE_opencv_gpu)
ocv_include_directories("${OpenCV_SOURCE_DIR}/modules/gpuarithm/include")
ocv_include_directories("${OpenCV_SOURCE_DIR}/modules/gpufilters/include")
ocv_include_directories("${OpenCV_SOURCE_DIR}/modules/gpu/include")
endif()

@ -2,7 +2,7 @@ SET(OPENCV_GPU_SAMPLES_REQUIRED_DEPS opencv_core opencv_flann opencv_imgproc ope
opencv_ml opencv_video opencv_objdetect opencv_features2d
opencv_calib3d opencv_legacy opencv_contrib opencv_gpu
opencv_nonfree opencv_softcascade opencv_superres
opencv_gpucodec opencv_gpuarithm)
opencv_gpucodec opencv_gpuarithm opencv_gpufilters)
ocv_check_dependencies(${OPENCV_GPU_SAMPLES_REQUIRED_DEPS})

Loading…
Cancel
Save