mirror of
https://github.com/opencv/opencv.git
synced 2024-12-15 09:49:13 +08:00
444 lines
13 KiB
C++
444 lines
13 KiB
C++
/*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.
|
|
//
|
|
//
|
|
// Intel License Agreement
|
|
// For Open Source Computer Vision Library
|
|
//
|
|
// Copyright (C) 2000, Intel Corporation, 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 Intel Corporation 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"
|
|
#include "opencv2/ts/ocl_test.hpp"
|
|
|
|
namespace opencv_test { namespace {
|
|
|
|
#define OCL_TUNING_MODE 0
|
|
#if OCL_TUNING_MODE
|
|
#define OCL_TUNING_MODE_ONLY(code) code
|
|
#else
|
|
#define OCL_TUNING_MODE_ONLY(code)
|
|
#endif
|
|
|
|
// image moments
|
|
class CV_MomentsTest : public cvtest::ArrayTest
|
|
{
|
|
public:
|
|
CV_MomentsTest(bool try_umat);
|
|
|
|
protected:
|
|
|
|
enum { MOMENT_COUNT = 25 };
|
|
int prepare_test_case( int test_case_idx );
|
|
void prepare_to_validation( int /*test_case_idx*/ );
|
|
void get_test_array_types_and_sizes( int test_case_idx, vector<vector<Size> >& sizes, vector<vector<int> >& types );
|
|
void get_minmax_bounds( int i, int j, int type, Scalar& low, Scalar& high );
|
|
double get_success_error_level( int test_case_idx, int i, int j );
|
|
void run_func();
|
|
bool is_binary;
|
|
bool try_umat_;
|
|
};
|
|
|
|
|
|
CV_MomentsTest::CV_MomentsTest(bool try_umat) :
|
|
try_umat_(try_umat)
|
|
{
|
|
test_array[INPUT].push_back(NULL);
|
|
test_array[OUTPUT].push_back(NULL);
|
|
test_array[REF_OUTPUT].push_back(NULL);
|
|
is_binary = false;
|
|
OCL_TUNING_MODE_ONLY(test_case_count = 10);
|
|
//element_wise_relative_error = false;
|
|
}
|
|
|
|
|
|
void CV_MomentsTest::get_minmax_bounds( int i, int j, int type, Scalar& low, Scalar& high )
|
|
{
|
|
cvtest::ArrayTest::get_minmax_bounds( i, j, type, low, high );
|
|
int depth = CV_MAT_DEPTH(type);
|
|
|
|
if( depth == CV_16U )
|
|
{
|
|
low = Scalar::all(0);
|
|
high = Scalar::all(1000);
|
|
}
|
|
else if( depth == CV_16S )
|
|
{
|
|
low = Scalar::all(-1000);
|
|
high = Scalar::all(1000);
|
|
}
|
|
else if( depth == CV_32F )
|
|
{
|
|
low = Scalar::all(-1);
|
|
high = Scalar::all(1);
|
|
}
|
|
}
|
|
|
|
void CV_MomentsTest::get_test_array_types_and_sizes( int test_case_idx,
|
|
vector<vector<Size> >& sizes, vector<vector<int> >& types )
|
|
{
|
|
RNG& rng = ts->get_rng();
|
|
cvtest::ArrayTest::get_test_array_types_and_sizes( test_case_idx, sizes, types );
|
|
int depth = cvtest::randInt(rng) % 4;
|
|
depth = depth == 0 ? CV_8U : depth == 1 ? CV_16U : depth == 2 ? CV_16S : CV_32F;
|
|
|
|
is_binary = cvtest::randInt(rng) % 2 != 0;
|
|
|
|
OCL_TUNING_MODE_ONLY(
|
|
depth = CV_8U;
|
|
is_binary = false;
|
|
sizes[INPUT][0] = Size(1024,768)
|
|
);
|
|
|
|
types[INPUT][0] = CV_MAKETYPE(depth, 1);
|
|
types[OUTPUT][0] = types[REF_OUTPUT][0] = CV_64FC1;
|
|
sizes[OUTPUT][0] = sizes[REF_OUTPUT][0] = cvSize(MOMENT_COUNT,1);
|
|
if(CV_MAT_DEPTH(types[INPUT][0])>=CV_32S)
|
|
sizes[INPUT][0].width = MAX(sizes[INPUT][0].width, 3);
|
|
|
|
cvmat_allowed = true;
|
|
}
|
|
|
|
|
|
double CV_MomentsTest::get_success_error_level( int /*test_case_idx*/, int /*i*/, int /*j*/ )
|
|
{
|
|
int depth = test_mat[INPUT][0].depth();
|
|
return depth != CV_32F ? FLT_EPSILON*10 : FLT_EPSILON*100;
|
|
}
|
|
|
|
int CV_MomentsTest::prepare_test_case( int test_case_idx )
|
|
{
|
|
int code = cvtest::ArrayTest::prepare_test_case( test_case_idx );
|
|
return code;
|
|
}
|
|
|
|
|
|
void CV_MomentsTest::run_func()
|
|
{
|
|
CvMoments* m = (CvMoments*)test_mat[OUTPUT][0].ptr<double>();
|
|
double* others = (double*)(m + 1);
|
|
if (try_umat_)
|
|
{
|
|
UMat u;
|
|
test_mat[INPUT][0].clone().copyTo(u);
|
|
OCL_TUNING_MODE_ONLY(
|
|
static double ttime = 0;
|
|
static int ncalls = 0;
|
|
moments(u, is_binary != 0);
|
|
double t = (double)getTickCount());
|
|
Moments new_m = moments(u, is_binary != 0);
|
|
OCL_TUNING_MODE_ONLY(
|
|
ttime += (double)getTickCount() - t;
|
|
ncalls++;
|
|
printf("%g\n", ttime/ncalls/u.total()));
|
|
*m = cvMoments(new_m);
|
|
}
|
|
else
|
|
cvMoments( test_array[INPUT][0], m, is_binary );
|
|
|
|
others[0] = cvGetNormalizedCentralMoment( m, 2, 0 );
|
|
others[1] = cvGetNormalizedCentralMoment( m, 1, 1 );
|
|
others[2] = cvGetNormalizedCentralMoment( m, 0, 2 );
|
|
others[3] = cvGetNormalizedCentralMoment( m, 3, 0 );
|
|
others[4] = cvGetNormalizedCentralMoment( m, 2, 1 );
|
|
others[5] = cvGetNormalizedCentralMoment( m, 1, 2 );
|
|
others[6] = cvGetNormalizedCentralMoment( m, 0, 3 );
|
|
}
|
|
|
|
|
|
void CV_MomentsTest::prepare_to_validation( int /*test_case_idx*/ )
|
|
{
|
|
Mat& src = test_mat[INPUT][0];
|
|
CvMoments m = cvMoments();
|
|
double* mdata = test_mat[REF_OUTPUT][0].ptr<double>();
|
|
int depth = src.depth();
|
|
int cn = src.channels();
|
|
int i, y, x, cols = src.cols;
|
|
double xc = 0., yc = 0.;
|
|
|
|
int coi = 0;
|
|
for( y = 0; y < src.rows; y++ )
|
|
{
|
|
double s0 = 0, s1 = 0, s2 = 0, s3 = 0;
|
|
uchar* ptr = src.ptr(y);
|
|
for( x = 0; x < cols; x++ )
|
|
{
|
|
double val;
|
|
if( depth == CV_8U )
|
|
val = ptr[x*cn + coi];
|
|
else if( depth == CV_16U )
|
|
val = ((ushort*)ptr)[x*cn + coi];
|
|
else if( depth == CV_16S )
|
|
val = ((short*)ptr)[x*cn + coi];
|
|
else
|
|
val = ((float*)ptr)[x*cn + coi];
|
|
|
|
if( is_binary )
|
|
val = val != 0;
|
|
|
|
s0 += val;
|
|
s1 += val*x;
|
|
s2 += val*x*x;
|
|
s3 += ((val*x)*x)*x;
|
|
}
|
|
|
|
m.m00 += s0;
|
|
m.m01 += s0*y;
|
|
m.m02 += (s0*y)*y;
|
|
m.m03 += ((s0*y)*y)*y;
|
|
|
|
m.m10 += s1;
|
|
m.m11 += s1*y;
|
|
m.m12 += (s1*y)*y;
|
|
|
|
m.m20 += s2;
|
|
m.m21 += s2*y;
|
|
|
|
m.m30 += s3;
|
|
}
|
|
|
|
if( m.m00 != 0 )
|
|
{
|
|
xc = m.m10/m.m00, yc = m.m01/m.m00;
|
|
m.inv_sqrt_m00 = 1./sqrt(fabs(m.m00));
|
|
}
|
|
|
|
for( y = 0; y < src.rows; y++ )
|
|
{
|
|
double s0 = 0, s1 = 0, s2 = 0, s3 = 0, y1 = y - yc;
|
|
uchar* ptr = src.ptr(y);
|
|
for( x = 0; x < cols; x++ )
|
|
{
|
|
double val, x1 = x - xc;
|
|
if( depth == CV_8U )
|
|
val = ptr[x*cn + coi];
|
|
else if( depth == CV_16U )
|
|
val = ((ushort*)ptr)[x*cn + coi];
|
|
else if( depth == CV_16S )
|
|
val = ((short*)ptr)[x*cn + coi];
|
|
else
|
|
val = ((float*)ptr)[x*cn + coi];
|
|
|
|
if( is_binary )
|
|
val = val != 0;
|
|
|
|
s0 += val;
|
|
s1 += val*x1;
|
|
s2 += val*x1*x1;
|
|
s3 += ((val*x1)*x1)*x1;
|
|
}
|
|
|
|
m.mu02 += s0*y1*y1;
|
|
m.mu03 += ((s0*y1)*y1)*y1;
|
|
|
|
m.mu11 += s1*y1;
|
|
m.mu12 += (s1*y1)*y1;
|
|
|
|
m.mu20 += s2;
|
|
m.mu21 += s2*y1;
|
|
|
|
m.mu30 += s3;
|
|
}
|
|
|
|
memcpy( mdata, &m, sizeof(m));
|
|
mdata += sizeof(m)/sizeof(m.m00);
|
|
|
|
/* calc normalized moments */
|
|
{
|
|
double inv_m00 = m.inv_sqrt_m00*m.inv_sqrt_m00;
|
|
double s2 = inv_m00*inv_m00; /* 1./(m00 ^ (2/2 + 1)) */
|
|
double s3 = s2*m.inv_sqrt_m00; /* 1./(m00 ^ (3/2 + 1)) */
|
|
|
|
mdata[0] = m.mu20 * s2;
|
|
mdata[1] = m.mu11 * s2;
|
|
mdata[2] = m.mu02 * s2;
|
|
|
|
mdata[3] = m.mu30 * s3;
|
|
mdata[4] = m.mu21 * s3;
|
|
mdata[5] = m.mu12 * s3;
|
|
mdata[6] = m.mu03 * s3;
|
|
}
|
|
|
|
double* a = test_mat[REF_OUTPUT][0].ptr<double>();
|
|
double* b = test_mat[OUTPUT][0].ptr<double>();
|
|
for( i = 0; i < MOMENT_COUNT; i++ )
|
|
{
|
|
if( fabs(a[i]) < 1e-3 )
|
|
a[i] = 0;
|
|
if( fabs(b[i]) < 1e-3 )
|
|
b[i] = 0;
|
|
}
|
|
}
|
|
|
|
|
|
// Hu invariants
|
|
class CV_HuMomentsTest : public cvtest::ArrayTest
|
|
{
|
|
public:
|
|
CV_HuMomentsTest();
|
|
|
|
protected:
|
|
|
|
enum { MOMENT_COUNT = 18, HU_MOMENT_COUNT = 7 };
|
|
|
|
int prepare_test_case( int test_case_idx );
|
|
void prepare_to_validation( int /*test_case_idx*/ );
|
|
void get_test_array_types_and_sizes( int test_case_idx, vector<vector<Size> >& sizes, vector<vector<int> >& types );
|
|
void get_minmax_bounds( int i, int j, int type, Scalar& low, Scalar& high );
|
|
double get_success_error_level( int test_case_idx, int i, int j );
|
|
void run_func();
|
|
};
|
|
|
|
|
|
CV_HuMomentsTest::CV_HuMomentsTest()
|
|
{
|
|
test_array[INPUT].push_back(NULL);
|
|
test_array[OUTPUT].push_back(NULL);
|
|
test_array[REF_OUTPUT].push_back(NULL);
|
|
}
|
|
|
|
|
|
void CV_HuMomentsTest::get_minmax_bounds( int i, int j, int type, Scalar& low, Scalar& high )
|
|
{
|
|
cvtest::ArrayTest::get_minmax_bounds( i, j, type, low, high );
|
|
low = Scalar::all(-10000);
|
|
high = Scalar::all(10000);
|
|
}
|
|
|
|
|
|
void CV_HuMomentsTest::get_test_array_types_and_sizes( int test_case_idx,
|
|
vector<vector<Size> >& sizes, vector<vector<int> >& types )
|
|
{
|
|
cvtest::ArrayTest::get_test_array_types_and_sizes( test_case_idx, sizes, types );
|
|
types[INPUT][0] = types[OUTPUT][0] = types[REF_OUTPUT][0] = CV_64FC1;
|
|
sizes[INPUT][0] = cvSize(MOMENT_COUNT,1);
|
|
sizes[OUTPUT][0] = sizes[REF_OUTPUT][0] = cvSize(HU_MOMENT_COUNT,1);
|
|
}
|
|
|
|
|
|
double CV_HuMomentsTest::get_success_error_level( int /*test_case_idx*/, int /*i*/, int /*j*/ )
|
|
{
|
|
return FLT_EPSILON;
|
|
}
|
|
|
|
|
|
|
|
int CV_HuMomentsTest::prepare_test_case( int test_case_idx )
|
|
{
|
|
int code = cvtest::ArrayTest::prepare_test_case( test_case_idx );
|
|
if( code > 0 )
|
|
{
|
|
// ...
|
|
}
|
|
|
|
return code;
|
|
}
|
|
|
|
|
|
void CV_HuMomentsTest::run_func()
|
|
{
|
|
cvGetHuMoments( test_mat[INPUT][0].ptr<CvMoments>(),
|
|
test_mat[OUTPUT][0].ptr<CvHuMoments>() );
|
|
}
|
|
|
|
|
|
void CV_HuMomentsTest::prepare_to_validation( int /*test_case_idx*/ )
|
|
{
|
|
CvMoments* m = test_mat[INPUT][0].ptr<CvMoments>();
|
|
CvHuMoments* hu = test_mat[REF_OUTPUT][0].ptr<CvHuMoments>();
|
|
|
|
double inv_m00 = m->inv_sqrt_m00*m->inv_sqrt_m00;
|
|
double s2 = inv_m00*inv_m00; /* 1./(m00 ^ (2/2 + 1)) */
|
|
double s3 = s2*m->inv_sqrt_m00; /* 1./(m00 ^ (3/2 + 1)) */
|
|
|
|
double nu20 = m->mu20 * s2;
|
|
double nu11 = m->mu11 * s2;
|
|
double nu02 = m->mu02 * s2;
|
|
|
|
double nu30 = m->mu30 * s3;
|
|
double nu21 = m->mu21 * s3;
|
|
double nu12 = m->mu12 * s3;
|
|
double nu03 = m->mu03 * s3;
|
|
|
|
#undef sqr
|
|
#define sqr(a) ((a)*(a))
|
|
|
|
hu->hu1 = nu20 + nu02;
|
|
hu->hu2 = sqr(nu20 - nu02) + 4*sqr(nu11);
|
|
hu->hu3 = sqr(nu30 - 3*nu12) + sqr(3*nu21 - nu03);
|
|
hu->hu4 = sqr(nu30 + nu12) + sqr(nu21 + nu03);
|
|
hu->hu5 = (nu30 - 3*nu12)*(nu30 + nu12)*(sqr(nu30 + nu12) - 3*sqr(nu21 + nu03)) +
|
|
(3*nu21 - nu03)*(nu21 + nu03)*(3*sqr(nu30 + nu12) - sqr(nu21 + nu03));
|
|
hu->hu6 = (nu20 - nu02)*(sqr(nu30 + nu12) - sqr(nu21 + nu03)) +
|
|
4*nu11*(nu30 + nu12)*(nu21 + nu03);
|
|
hu->hu7 = (3*nu21 - nu03)*(nu30 + nu12)*(sqr(nu30 + nu12) - 3*sqr(nu21 + nu03)) +
|
|
(3*nu12 - nu30)*(nu21 + nu03)*(3*sqr(nu30 + nu12) - sqr(nu21 + nu03));
|
|
}
|
|
|
|
|
|
TEST(Imgproc_Moments, accuracy) { CV_MomentsTest test(false); test.safe_run(); }
|
|
OCL_TEST(Imgproc_Moments, accuracy) { CV_MomentsTest test(true); test.safe_run(); }
|
|
TEST(Imgproc_HuMoments, accuracy) { CV_HuMomentsTest test; test.safe_run(); }
|
|
|
|
class CV_SmallContourMomentTest : public cvtest::BaseTest
|
|
{
|
|
public:
|
|
CV_SmallContourMomentTest() {}
|
|
~CV_SmallContourMomentTest() {}
|
|
protected:
|
|
void run(int)
|
|
{
|
|
try
|
|
{
|
|
vector<Point> points;
|
|
points.push_back(Point(50, 56));
|
|
points.push_back(Point(53, 53));
|
|
points.push_back(Point(46, 54));
|
|
points.push_back(Point(49, 51));
|
|
|
|
Moments m = moments(points, false);
|
|
double area = contourArea(points);
|
|
|
|
CV_Assert( m.m00 == 0 && m.m01 == 0 && m.m10 == 0 && area == 0 );
|
|
}
|
|
catch(...)
|
|
{
|
|
ts->set_failed_test_info(cvtest::TS::FAIL_MISMATCH);
|
|
}
|
|
}
|
|
};
|
|
|
|
TEST(Imgproc_ContourMoment, small) { CV_SmallContourMomentTest test; test.safe_run(); }
|
|
|
|
}} // namespace
|