opencv/modules/photo/perf/perf_cuda.cpp

194 lines
5.9 KiB
C++
Raw Normal View History

/*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"
2013-07-24 16:01:34 +08:00
#include "opencv2/photo/cuda.hpp"
#include "opencv2/ts/cuda_perf.hpp"
#include "opencv2/opencv_modules.hpp"
2013-07-23 20:59:34 +08:00
#if defined (HAVE_CUDA) && defined(HAVE_OPENCV_CUDAARITHM) && defined(HAVE_OPENCV_CUDAIMGPROC)
using namespace std;
using namespace testing;
using namespace perf;
2013-07-24 17:55:18 +08:00
#define CUDA_DENOISING_IMAGE_SIZES testing::Values(perf::szVGA, perf::sz720p)
//////////////////////////////////////////////////////////////////////
// nonLocalMeans
2012-10-08 17:58:03 +08:00
DEF_PARAM_TEST(Sz_Depth_Cn_WinSz_BlockSz, cv::Size, MatDepth, MatCn, int, int);
2013-07-24 17:55:18 +08:00
PERF_TEST_P(Sz_Depth_Cn_WinSz_BlockSz, CUDA_NonLocalMeans,
Combine(CUDA_DENOISING_IMAGE_SIZES,
Values<MatDepth>(CV_8U),
2013-07-24 17:55:18 +08:00
CUDA_CHANNELS_1_3,
Values(21),
Values(5)))
{
2013-03-20 15:49:33 +08:00
declare.time(600.0);
const cv::Size size = GET_PARAM(0);
const int depth = GET_PARAM(1);
const int channels = GET_PARAM(2);
const int search_widow_size = GET_PARAM(3);
const int block_size = GET_PARAM(4);
const float h = 10;
const int borderMode = cv::BORDER_REFLECT101;
const int type = CV_MAKE_TYPE(depth, channels);
cv::Mat src(size, type);
declare.in(src, WARMUP_RNG);
2013-07-24 17:55:18 +08:00
if (PERF_RUN_CUDA())
{
2013-08-28 19:45:13 +08:00
const cv::cuda::GpuMat d_src(src);
cv::cuda::GpuMat dst;
2013-08-28 19:45:13 +08:00
TEST_CYCLE() cv::cuda::nonLocalMeans(d_src, dst, h, search_widow_size, block_size, borderMode);
2013-07-24 17:55:18 +08:00
CUDA_SANITY_CHECK(dst);
}
else
{
FAIL_NO_CPU();
}
2012-09-27 22:11:06 +08:00
}
//////////////////////////////////////////////////////////////////////
// fastNonLocalMeans
2012-10-08 17:58:03 +08:00
DEF_PARAM_TEST(Sz_Depth_Cn_WinSz_BlockSz, cv::Size, MatDepth, MatCn, int, int);
2012-09-27 22:11:06 +08:00
2013-07-24 17:55:18 +08:00
PERF_TEST_P(Sz_Depth_Cn_WinSz_BlockSz, CUDA_FastNonLocalMeans,
Combine(CUDA_DENOISING_IMAGE_SIZES,
Values<MatDepth>(CV_8U),
2013-07-24 17:55:18 +08:00
CUDA_CHANNELS_1_3,
Values(21),
Values(7)))
2012-09-27 22:11:06 +08:00
{
declare.time(60.0);
const cv::Size size = GET_PARAM(0);
const int depth = GET_PARAM(1);
const int search_widow_size = GET_PARAM(2);
const int block_size = GET_PARAM(3);
const float h = 10;
const int type = CV_MAKE_TYPE(depth, 1);
2012-10-04 23:36:48 +08:00
cv::Mat src(size, type);
declare.in(src, WARMUP_RNG);
2012-09-27 22:11:06 +08:00
2013-07-24 17:55:18 +08:00
if (PERF_RUN_CUDA())
2012-10-04 23:36:48 +08:00
{
2013-08-28 19:45:13 +08:00
cv::cuda::FastNonLocalMeansDenoising fnlmd;
2012-10-04 23:36:48 +08:00
2013-08-28 19:45:13 +08:00
const cv::cuda::GpuMat d_src(src);
cv::cuda::GpuMat dst;
2012-10-04 23:36:48 +08:00
TEST_CYCLE() fnlmd.simpleMethod(d_src, dst, h, search_widow_size, block_size);
2013-07-24 17:55:18 +08:00
CUDA_SANITY_CHECK(dst);
2012-10-04 23:36:48 +08:00
}
else
{
cv::Mat dst;
TEST_CYCLE() cv::fastNlMeansDenoising(src, dst, h, block_size, search_widow_size);
CPU_SANITY_CHECK(dst);
2012-10-04 23:36:48 +08:00
}
}
//////////////////////////////////////////////////////////////////////
// fastNonLocalMeans (colored)
2012-10-08 17:58:03 +08:00
DEF_PARAM_TEST(Sz_Depth_WinSz_BlockSz, cv::Size, MatDepth, int, int);
2012-10-04 23:36:48 +08:00
2013-07-24 17:55:18 +08:00
PERF_TEST_P(Sz_Depth_WinSz_BlockSz, CUDA_FastNonLocalMeansColored,
Combine(CUDA_DENOISING_IMAGE_SIZES,
Values<MatDepth>(CV_8U),
Values(21),
Values(7)))
2012-10-04 23:36:48 +08:00
{
declare.time(60.0);
const cv::Size size = GET_PARAM(0);
const int depth = GET_PARAM(1);
const int search_widow_size = GET_PARAM(2);
const int block_size = GET_PARAM(3);
2012-09-27 22:11:06 +08:00
const float h = 10;
const int type = CV_MAKE_TYPE(depth, 3);
2012-09-27 22:11:06 +08:00
cv::Mat src(size, type);
declare.in(src, WARMUP_RNG);
2012-09-27 22:11:06 +08:00
2013-07-24 17:55:18 +08:00
if (PERF_RUN_CUDA())
2012-09-27 22:11:06 +08:00
{
2013-08-28 19:45:13 +08:00
cv::cuda::FastNonLocalMeansDenoising fnlmd;
2012-10-04 23:36:48 +08:00
2013-08-28 19:45:13 +08:00
const cv::cuda::GpuMat d_src(src);
cv::cuda::GpuMat dst;
2012-09-27 22:11:06 +08:00
TEST_CYCLE() fnlmd.labMethod(d_src, dst, h, h, search_widow_size, block_size);
2013-07-24 17:55:18 +08:00
CUDA_SANITY_CHECK(dst);
2012-09-27 22:11:06 +08:00
}
else
{
cv::Mat dst;
TEST_CYCLE() cv::fastNlMeansDenoisingColored(src, dst, h, h, block_size, search_widow_size);
CPU_SANITY_CHECK(dst);
2012-09-27 22:11:06 +08:00
}
}
#endif