Add perfomance test

diff --git a/modules/video/perf/opencl/perf_optflow_farneback.cpp b/modules/video/perf/opencl/perf_optflow_farneback.cpp
new file mode 100644 (file)
index 0000000..2b968aa
--- /dev/null
+++ b/modules/video/perf/opencl/perf_optflow_farneback.cpp
@@ -0,0 +1,114 @@
+/*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) 2010-2012, Multicoreware, Inc., all rights reserved.
+// Copyright (C) 2010-2012, Advanced Micro Devices, Inc., all rights reserved.
+// Third party copyrights are property of their respective owners.
+//
+// @Authors
+//    Fangfang Bai, fangfang@multicorewareinc.com
+//    Jin Ma,       jin@multicorewareinc.com
+//
+// 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"
+#include "opencv2/ts/ocl_perf.hpp"
+
+using std::tr1::make_tuple;
+
+#ifdef HAVE_OPENCL
+
+namespace cvtest {
+namespace ocl {
+
+///////////// FarnebackOpticalFlow ////////////////////////
+CV_ENUM(farneFlagType, 0, OPTFLOW_FARNEBACK_GAUSSIAN)
+
+typedef tuple< tuple<int, double>, farneFlagType, bool > FarnebackOpticalFlowParams;
+typedef TestBaseWithParam<FarnebackOpticalFlowParams> FarnebackOpticalFlowFixture;
+
+OCL_PERF_TEST_P(FarnebackOpticalFlowFixture, FarnebackOpticalFlow,
+                ::testing::Combine(
+                    ::testing::Values(
+                                      make_tuple<int, double>(5, 1.1),
+                                      make_tuple<int, double>(7, 1.5)
+                                     ),
+                    farneFlagType::all(),
+                    ::testing::Bool()
+                    )
+                )
+{
+    Mat frame0 = imread(getDataPath("gpu/opticalflow/rubberwhale1.png"), cv::IMREAD_GRAYSCALE);
+    ASSERT_FALSE(frame0.empty()) << "can't load rubberwhale1.png";
+
+    Mat frame1 = imread(getDataPath("gpu/opticalflow/rubberwhale2.png"), cv::IMREAD_GRAYSCALE);
+    ASSERT_FALSE(frame1.empty()) << "can't load rubberwhale2.png";
+
+    const Size srcSize = frame0.size();
+
+    const int numLevels = 5;
+    const int winSize = 13;
+    const int numIters = 10;
+
+    const FarnebackOpticalFlowParams params = GetParam();
+    const tuple<int, double> polyParams = get<0>(params);
+    const int polyN = get<0>(polyParams);
+    const double polySigma = get<1>(polyParams);
+    const double pyrScale = 0.5;
+    int flags = get<1>(params);
+    const bool useInitFlow = get<2>(params);
+    const double eps = 1.5;
+
+    UMat uFrame0; frame0.copyTo(uFrame0);
+    UMat uFrame1; frame1.copyTo(uFrame1);
+    UMat uFlow(srcSize, CV_32FC2);
+    declare.in(uFrame0, uFrame1, WARMUP_READ).out(uFlow, WARMUP_READ);
+    if (useInitFlow)
+    {
+        cv::calcOpticalFlowFarneback(uFrame0, uFrame1, uFlow, pyrScale, numLevels, winSize, numIters, polyN, polySigma, flags);
+        flags |= OPTFLOW_USE_INITIAL_FLOW;
+    }
+
+    OCL_TEST_CYCLE()
+            cv::calcOpticalFlowFarneback(uFrame0, uFrame1, uFlow, pyrScale, numLevels, winSize, numIters, polyN, polySigma, flags);
+
+
+    SANITY_CHECK(uFlow, eps);
+}
+
+} } // namespace cvtest::ocl
+
+#endif // HAVE_OPENCL
\ No newline at end of file

diff --git a/modules/video/src/optflowgf.cpp b/modules/video/src/optflowgf.cpp
index b237e54..73aedec 100644 (file)
--- a/modules/video/src/optflowgf.cpp
+++ b/modules/video/src/optflowgf.cpp
@@ -895,10 +895,14 @@ private:
 #endif
         size_t globalsize[2] = { DIVUP(src.cols, localsize[0] - 2*polyN) * localsize[0], src.rows};
 
+#if 0
         const cv::ocl::Device &device = cv::ocl::Device::getDefault();
         bool useDouble = (0 != device.doubleFPConfig());
 
         cv::String build_options = cv::format("-D polyN=%d -D USE_DOUBLE=%d", polyN, useDouble ? 1 : 0);
+#else
+        cv::String build_options = cv::format("-D polyN=%d", polyN);
+#endif
         ocl::Kernel kernel;
         if (!kernel.create("polynomialExpansion", cv::ocl::video::optical_flow_farneback_oclsrc, build_options))
             return false;
@@ -915,9 +919,11 @@ private:
         idxArg = kernel.set(idxArg, ocl::KernelArg::PtrReadOnly(m_xg));
         idxArg = kernel.set(idxArg, ocl::KernelArg::PtrReadOnly(m_xxg));
         idxArg = kernel.set(idxArg, (void *)NULL, smem_size);
+#if 0
         if (useDouble)
             idxArg = kernel.set(idxArg, (void *)m_igd, 4 * sizeof(double));
         else
+#endif
             idxArg = kernel.set(idxArg, (void *)m_ig, 4 * sizeof(float));
         return kernel.run(2, globalsize, localsize, false);
     }