opencv/modules/ocl/test/test_kmeans.cpp

236 lines
8.1 KiB
C++
Raw Normal View History

2013-07-03 13:13:04 +08:00
/*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
// Erping Pang, pang_er_ping@163.com
// Xiaopeng Fu, fuxiaopeng2222@163.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
2013-10-25 22:00:46 +08:00
// and/or other materials provided with the distribution.
2013-07-03 13:13:04 +08:00
//
// * 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"
2013-07-03 13:13:04 +08:00
#ifdef HAVE_OPENCL
using namespace cvtest;
using namespace testing;
using namespace std;
using namespace cv;
#define OCL_KMEANS_USE_INITIAL_LABELS 1
#define OCL_KMEANS_PP_CENTERS 2
PARAM_TEST_CASE(Kmeans, int, int, int)
{
int type;
int K;
int flags;
2013-11-06 00:03:49 +08:00
Mat src ;
2013-07-03 13:13:04 +08:00
ocl::oclMat d_src, d_dists;
Mat labels, centers;
ocl::oclMat d_labels, d_centers;
2013-10-03 00:18:48 +08:00
virtual void SetUp()
{
2013-07-03 13:13:04 +08:00
K = GET_PARAM(0);
type = GET_PARAM(1);
flags = GET_PARAM(2);
// MWIDTH=256, MHEIGHT=256. defined in utility.hpp
2013-11-06 00:03:49 +08:00
Size size = Size(MWIDTH, MHEIGHT);
2013-07-03 13:13:04 +08:00
src.create(size, type);
int row_idx = 0;
const int max_neighbour = MHEIGHT / K - 1;
CV_Assert(K <= MWIDTH);
for(int i = 0; i < K; i++ )
{
Mat center_row_header = src.row(row_idx);
center_row_header.setTo(0);
int nchannel = center_row_header.channels();
for(int j = 0; j < nchannel; j++)
center_row_header.at<float>(0, i*nchannel+j) = 50000.0;
2013-08-21 20:44:09 +08:00
for(int j = 0; (j < max_neighbour) ||
2013-07-03 13:13:04 +08:00
(i == K-1 && j < max_neighbour + MHEIGHT%K); j ++)
{
Mat cur_row_header = src.row(row_idx + 1 + j);
center_row_header.copyTo(cur_row_header);
2013-10-03 00:18:48 +08:00
Mat tmpmat = randomMat(cur_row_header.size(), cur_row_header.type(), -200, 200, false);
2013-07-03 13:13:04 +08:00
cur_row_header += tmpmat;
}
row_idx += 1 + max_neighbour;
}
}
};
OCL_TEST_P(Kmeans, Mat){
2013-07-03 13:13:04 +08:00
if(flags & KMEANS_USE_INITIAL_LABELS)
{
// inital a given labels
labels.create(src.rows, 1, CV_32S);
int *label = labels.ptr<int>();
for(int i = 0; i < src.rows; i++)
label[i] = rng.uniform(0, K);
d_labels.upload(labels);
}
d_src.upload(src);
for(int j = 0; j < LOOP_TIMES; j++)
{
kmeans(src, K, labels,
TermCriteria( TermCriteria::EPS + TermCriteria::MAX_ITER, 100, 0),
2013-07-03 13:13:04 +08:00
1, flags, centers);
ocl::kmeans(d_src, K, d_labels,
TermCriteria( TermCriteria::EPS + TermCriteria::MAX_ITER, 100, 0),
2013-07-03 13:13:04 +08:00
1, flags, d_centers);
Mat dd_labels(d_labels);
Mat dd_centers(d_centers);
if(flags & KMEANS_USE_INITIAL_LABELS)
{
EXPECT_MAT_NEAR(labels, dd_labels, 0);
EXPECT_MAT_NEAR(centers, dd_centers, 1e-3);
2013-08-21 20:44:09 +08:00
}
else
2013-07-03 13:13:04 +08:00
{
int row_idx = 0;
for(int i = 0; i < K; i++)
{
// verify lables with ground truth resutls
int label = labels.at<int>(row_idx);
int header_label = dd_labels.at<int>(row_idx);
for(int j = 0; (j < MHEIGHT/K)||(i == K-1 && j < MHEIGHT/K+MHEIGHT%K); j++)
{
ASSERT_NEAR(labels.at<int>(row_idx+j), label, 0);
ASSERT_NEAR(dd_labels.at<int>(row_idx+j), header_label, 0);
}
// verify centers
float *center = centers.ptr<float>(label);
float *header_center = dd_centers.ptr<float>(header_label);
for(int t = 0; t < centers.cols; t++)
ASSERT_NEAR(center[t], header_center[t], 1e-3);
row_idx += MHEIGHT/K;
}
}
}
}
2013-07-03 13:13:04 +08:00
INSTANTIATE_TEST_CASE_P(OCL_ML, Kmeans, Combine(
Values(3, 5, 8),
Values(CV_32FC1, CV_32FC2, CV_32FC4),
2013-08-21 20:44:09 +08:00
Values(OCL_KMEANS_USE_INITIAL_LABELS/*, OCL_KMEANS_PP_CENTERS*/)));
2013-07-03 13:13:04 +08:00
/////////////////////////////// DistanceToCenters //////////////////////////////////////////
2013-11-06 00:03:49 +08:00
CV_ENUM(DistType, NORM_L1, NORM_L2SQR)
PARAM_TEST_CASE(distanceToCenters, DistType, bool)
{
int distType;
bool useRoi;
2013-11-06 00:03:49 +08:00
Mat src, centers, src_roi, centers_roi;
ocl::oclMat ocl_src, ocl_centers, ocl_src_roi, ocl_centers_roi;
virtual void SetUp()
{
distType = GET_PARAM(0);
useRoi = GET_PARAM(1);
}
void random_roi()
{
2013-11-06 00:03:49 +08:00
Size roiSizeSrc = randomSize(1, MAX_VALUE);
Size roiSizeCenters = randomSize(1, MAX_VALUE);
roiSizeSrc.width = roiSizeCenters.width;
2013-11-06 00:03:49 +08:00
Border srcBorder = randomBorder(0, useRoi ? MAX_VALUE : 0);
randomSubMat(src, src_roi, roiSizeSrc, srcBorder, CV_32FC1, -MAX_VALUE, MAX_VALUE);
Border centersBorder = randomBorder(0, useRoi ? 500 : 0);
2013-11-06 00:03:49 +08:00
randomSubMat(centers, centers_roi, roiSizeCenters, centersBorder, CV_32FC1, -MAX_VALUE, MAX_VALUE);
2013-11-06 00:03:49 +08:00
for (int i = 0; i < centers.rows; i++)
centers.at<float>(i, randomInt(0, centers.cols)) = (float)randomDouble(SHRT_MAX, INT_MAX);
2013-11-06 00:03:49 +08:00
generateOclMat(ocl_src, ocl_src_roi, src, roiSizeSrc, srcBorder);
generateOclMat(ocl_centers, ocl_centers_roi, centers, roiSizeCenters, centersBorder);
}
};
OCL_TEST_P(distanceToCenters, Accuracy)
{
2013-11-06 00:03:49 +08:00
for (int j = 0; j < LOOP_TIMES; j++)
{
random_roi();
Mat labels, dists;
2013-11-06 00:03:49 +08:00
ocl::distanceToCenters(ocl_src_roi, ocl_centers_roi, dists, labels, distType);
2013-11-06 00:03:49 +08:00
EXPECT_EQ(dists.size(), labels.size());
Mat batch_dists;
cv::batchDistance(src_roi, centers_roi, batch_dists, CV_32FC1, noArray(), distType);
2013-11-06 00:03:49 +08:00
std::vector<float> gold_dists_v;
gold_dists_v.reserve(batch_dists.rows);
2013-11-06 00:03:49 +08:00
for (int i = 0; i < batch_dists.rows; i++)
{
Mat r = batch_dists.row(i);
double mVal;
Point mLoc;
minMaxLoc(r, &mVal, NULL, &mLoc, NULL);
2013-11-06 00:03:49 +08:00
int ocl_label = labels.at<int>(i, 0);
EXPECT_EQ(mLoc.x, ocl_label);
2013-11-06 00:03:49 +08:00
gold_dists_v.push_back(static_cast<float>(mVal));
}
2013-11-06 00:03:49 +08:00
double relative_error = cv::norm(Mat(gold_dists_v), dists, NORM_INF | NORM_RELATIVE);
ASSERT_LE(relative_error, 1e-5);
}
}
2013-11-06 00:03:49 +08:00
INSTANTIATE_TEST_CASE_P (OCL_ML, distanceToCenters, Combine(DistType::all(), Bool()));
2013-07-03 13:13:04 +08:00
#endif