TEST_P(EstimateAffinePartial2D, test2Points)
{
// try more transformations
for (size_t i = 0; i < 500; ++i)
{
Mat aff = rngPartialAffMat();
// setting points that are no in the same line
Mat fpts(1, 2, CV_32FC2);
Mat tpts(1, 2, CV_32FC2);
fpts.at<Point2f>(0) = Point2f( rngIn(1,2), rngIn(5,6) );
fpts.at<Point2f>(1) = Point2f( rngIn(3,4), rngIn(3,4) );
transform(fpts, tpts, aff);
vector<uchar> inliers;
Mat aff_est = estimateAffinePartial2D(fpts, tpts, inliers, GetParam() /* method */);
EXPECT_NEAR(0., cvtest::norm(aff_est, aff, NORM_INF), 1e-3);
// all must be inliers
EXPECT_EQ(countNonZero(inliers), 2);
}
}
// test conversion from other datatypes than float
TEST_P(EstimateAffinePartial2D, testConversion)
{
Mat aff = rngPartialAffMat();
aff.convertTo(aff, CV_32S); // convert to int to transform ints properly
std::vector<Point> fpts(3);
std::vector<Point> tpts(3);
fpts[0] = Point2f( rngIn(1,2), rngIn(5,6) );
fpts[1] = Point2f( rngIn(3,4), rngIn(3,4) );
fpts[2] = Point2f( rngIn(1,2), rngIn(3,4) );
transform(fpts, tpts, aff);
vector<uchar> inliers;
Mat aff_est = estimateAffinePartial2D(fpts, tpts, inliers, GetParam() /* method */);
ASSERT_FALSE(aff_est.empty());
aff.convertTo(aff, CV_64F); // need to convert back before compare
EXPECT_NEAR(0., cvtest::norm(aff_est, aff, NORM_INF), 1e-3);
// all must be inliers
EXPECT_EQ(countNonZero(inliers), 3);
}
PERF_TEST_P( EstimateAffine, EstimateAffinePartial2D, ESTIMATE_PARAMS )
{
AffineParams params = GetParam();
const int n = get<0>(params);
const double confidence = get<1>(params);
const int method = get<2>(params);
const size_t refining = get<3>(params);
Mat aff = rngPartialAffMat();
int m;
// LMEDS can't handle more than 50% outliers (by design)
if (method == LMEDS)
m = 3*n/5;
else
m = 2*n/5;
const float shift_outl = 15.f; const float noise_level = 20.f;
Mat fpts(1, n, CV_32FC2);
Mat tpts(1, n, CV_32FC2);
randu(fpts, 0., 100.);
transform(fpts, tpts, aff);
/* adding noise*/
Mat outliers = tpts.colRange(m, n);
outliers.reshape(1) += shift_outl;
Mat noise (outliers.size(), outliers.type());
randu(noise, 0., noise_level);
outliers += noise;
Mat aff_est;
vector<uchar> inliers (n);
warmup(inliers, WARMUP_WRITE);
warmup(fpts, WARMUP_READ);
warmup(tpts, WARMUP_READ);
TEST_CYCLE()
{
aff_est = estimateAffinePartial2D(fpts, tpts, inliers, method, 3, 2000, confidence, refining);
}
// we already have accuracy tests
SANITY_CHECK_NOTHING();
}
TEST_P(EstimateAffinePartial2D, testNPoints)
{
// try more transformations
for (size_t i = 0; i < 500; ++i)
{
Mat aff = rngPartialAffMat();
const int method = GetParam();
const int n = 100;
int m;
// LMEDS can't handle more than 50% outliers (by design)
if (method == LMEDS)
m = 3*n/5;
else
m = 2*n/5;
const float shift_outl = 15.f;
const float noise_level = 20.f;
Mat fpts(1, n, CV_32FC2);
Mat tpts(1, n, CV_32FC2);
randu(fpts, 0., 100.);
transform(fpts, tpts, aff);
/* adding noise to some points */
Mat outliers = tpts.colRange(m, n);
outliers.reshape(1) += shift_outl;
Mat noise (outliers.size(), outliers.type());
randu(noise, 0., noise_level);
outliers += noise;
vector<uchar> inliers;
Mat aff_est = estimateAffinePartial2D(fpts, tpts, inliers, method);
EXPECT_FALSE(aff_est.empty());
EXPECT_NEAR(0., cvtest::norm(aff_est, aff, NORM_INF), 1e-4);
bool inliers_good = count(inliers.begin(), inliers.end(), 1) == m &&
m == accumulate(inliers.begin(), inliers.begin() + m, 0);
EXPECT_TRUE(inliers_good);
}
}