Fixed bugs in BruteForceMatcher and its specialization

This commit is contained in:
Ilya Lysenkov
2010-08-04 08:26:30 +00:00
parent 4e60decad3
commit 20e407b235
3 changed files with 111 additions and 25 deletions

View File

@@ -427,8 +427,11 @@ Ptr<DescriptorMatcher> createDescriptorMatcher( const string& descriptorMatcherT
* BruteForceMatcher L2 specialization *
\****************************************************************************************/
template<>
void BruteForceMatcher<L2<float> >::matchImpl( const Mat& query, const Mat& /*mask*/, vector<int>& matches ) const
void BruteForceMatcher<L2<float> >::matchImpl( const Mat& query, const Mat& mask, vector<DMatch>& matches ) const
{
assert( mask.empty() || (mask.rows == query.rows && mask.cols == train.rows) );
assert( query.cols == train.cols || query.empty() || train.empty() );
matches.clear();
matches.reserve( query.rows );
#if (!defined HAVE_EIGEN2)
@@ -440,9 +443,27 @@ void BruteForceMatcher<L2<float> >::matchImpl( const Mat& query, const Mat& /*ma
{
Mat distances = (-2)*query.row(i)*desc_2t;
distances += norms;
DMatch match;
match.indexTrain = -1;
double minVal;
Point minLoc;
minMaxLoc ( distances, 0, 0, &minLoc );
matches.push_back( minLoc.x );
if( mask.empty() )
{
minMaxLoc ( distances, &minVal, 0, &minLoc );
}
else
{
minMaxLoc ( distances, &minVal, 0, &minLoc, 0, mask.row( i ) );
}
match.indexTrain = minLoc.x;
if( match.indexTrain != -1 )
{
match.indexQuery = i;
double queryNorm = norm( query.row(i) );
match.distance = sqrt( minVal + queryNorm*queryNorm );
matches.push_back( match );
}
}
#else
@@ -451,21 +472,46 @@ void BruteForceMatcher<L2<float> >::matchImpl( const Mat& query, const Mat& /*ma
cv2eigen( query.t(), desc1t);
cv2eigen( train, desc2 );
//Eigen::Matrix<float, Eigen::Dynamic, 1> norms = desc2.rowwise().squaredNorm();
Eigen::Matrix<float, Eigen::Dynamic, 1> norms = desc2.rowwise().squaredNorm() / 2;
for( int i=0;i<query.rows;i++ )
if( mask.empty() )
{
//Eigen::Matrix<float, Eigen::Dynamic, 1> distances = (-2) * (desc2*desc1t.col(i));
Eigen::Matrix<float, Eigen::Dynamic, 1> distances = desc2*desc1t.col(i);
for( int i=0;i<query.rows;i++ )
{
Eigen::Matrix<float, Eigen::Dynamic, 1> distances = desc2*desc1t.col(i);
distances -= norms;
DMatch match;
match.indexQuery = i;
match.distance = sqrt( (-2)*distances.maxCoeff( &match.indexTrain ) + desc1t.col(i).squaredNorm() );
matches.push_back( match );
}
}
else
{
for( int i=0;i<query.rows;i++ )
{
Eigen::Matrix<float, Eigen::Dynamic, 1> distances = desc2*desc1t.col(i);
distances -= norms;
//distances += norms;
distances -= norms;
float maxCoeff = -std::numeric_limits<float>::max();
DMatch match;
match.indexTrain = -1;
for( int j=0;j<train.rows;j++ )
{
if( possibleMatch( mask, i, j ) && distances( j, 0 ) > maxCoeff )
{
maxCoeff = distances( j, 0 );
match.indexTrain = j;
}
}
int idx;
//distances.minCoeff(&idx);
distances.maxCoeff(&idx);
matches.push_back( idx );
if( match.indexTrain != -1 )
{
match.indexQuery = i;
match.distance = sqrt( (-2)*maxCoeff + desc1t.col(i).squaredNorm() );
matches.push_back( match );
}
}
}
#endif
}