generic-library/opencv
1
0
Fork 0
Code Issues Pull Requests Releases Wiki Activity

changed the dot

Browse Source
This commit is contained in:
Maria Dimashova
2011-05-10 12:01:42 +00:00
parent d0a91f8f19
commit f4c74eb532
3 changed files with 187 additions and 117 deletions
Download Patch File Download Diff File Expand all files Collapse all files

212
modules/objdetect/src/dotdetector.cpp
View File

@@ -56,7 +56,7 @@ static void readDirContent( const string& descrFilename, vector<string>& names )
{
names.clear();
ifstream file( descrFilename.c_str() );
ifstream file( descrFilename.c_str(), ifstream::in );
if ( !file.is_open() )
return;
@@ -233,7 +233,7 @@ static void quantizeToTrain( const Mat& _magnitudesExt, const Mat& _anglesExt, c
subMagnitudes.copyTo( subMagnitudesCopy );
Mat subAngles( anglesExt, shiftedRect );
double maxMagnitude;
double maxMagnitude = 0;
int strongestCount = 0;
for( ; strongestCount < params.maxStrongestCount; strongestCount++ )
{
@@ -268,16 +268,17 @@ static void quantizeToTrain( const Mat& _magnitudesExt, const Mat& _anglesExt, c
}
}
static void quantizeToDetect( const Mat& _magnitudes, const Mat& angles, Mat& quantizedImage, const DOTDetector::TrainParams& params )
static void quantizeToDetect( const Mat& _magnitudes, const Mat& angles,
Mat& quantizedImage, int regionSize, const DOTDetector::TrainParams& params )
{
Mat magnitudes; _magnitudes.copyTo( magnitudes );
const int verticalRegionCount = magnitudes.rows / params.regionSize;
const int horizontalRegionCount = magnitudes.cols / params.regionSize;
const int verticalRegionCount = magnitudes.rows / regionSize;
const int horizontalRegionCount = magnitudes.cols / regionSize;
quantizedImage = Mat( verticalRegionCount, horizontalRegionCount, CV_8UC1, Scalar::all(0) );
Rect curRect(0, 0, params.regionSize, params.regionSize);
Rect curRect(0, 0, regionSize, regionSize);
const int maxStrongestCount = 1;
for( int vRegIdx = 0; vRegIdx < verticalRegionCount; vRegIdx++ )
{
@@ -311,10 +312,10 @@ static void quantizeToDetect( const Mat& _magnitudes, const Mat& angles, Mat& qu
curRectBits |= 1 << DOTDetector::TrainParams::BIN_COUNT;
quantizedImage.at<uchar>(vRegIdx, hRegIdx) = curRectBits;
curRect.x += params.regionSize;
curRect.x += regionSize;
}
curRect.x = 0;
curRect.y += params.regionSize;
curRect.y += regionSize;
}
}
@@ -327,7 +328,7 @@ inline void andQuantizedImages( const Mat& queryQuantizedImage, const Mat& train
int area = cv::countNonZero( trainQuantizedImage );
ratio = (float)nonZeroCount / area;
texturelessRatio = texturelessCount / nonZeroCount;
texturelessRatio = (float)texturelessCount / nonZeroCount;
}
static void computeTrainUsedStrongestMask( const Mat& _magnitudesExt, const Mat& _anglesExt, const Mat& maskExt, const Mat& quantizedImage,
@@ -402,10 +403,10 @@ DOTDetector::TrainParams::TrainParams( const Size& _winSize, int _regionSize, in
maxStrongestCount(_maxStrongestCount), maxNonzeroBits(_maxNonzeroBits),
minRatio(_minRatio)
{
asserts();
isConsistent();
}
void DOTDetector::TrainParams::asserts() const
void DOTDetector::TrainParams::isConsistent() const
{
CV_Assert( winSize.width > 0 && winSize.height > 0 );
CV_Assert( regionSize > 0 && regionSize % 2 == 1);
@@ -433,7 +434,7 @@ void DOTDetector::TrainParams::read( FileNode& fn )
minRatio = fn["minRatio"];
asserts();
isConsistent();
}
void DOTDetector::TrainParams::write( FileStorage& fs ) const
@@ -459,10 +460,10 @@ DOTDetector::DetectParams::DetectParams( float _minRatio, int _minRegionSize, in
minRatio(_minRatio), minRegionSize(_minRegionSize), maxRegionSize(_maxRegionSize), regionSizeStep(_regionSizeStep),
isGroup(_isGroup), groupThreshold(_groupThreshold), groupEps(_groupEps)
{
asserts();
isConsistent();
}
void DOTDetector::DetectParams::asserts( float minTrainRatio ) const
void DOTDetector::DetectParams::isConsistent( float minTrainRatio ) const
{
CV_Assert( minRatio > 0 && minRatio < 1 );
CV_Assert( minRatio <= minTrainRatio );
@@ -484,22 +485,30 @@ void DOTDetector::DetectParams::asserts( float minTrainRatio ) const
* DOTDetector::DOTTemplate
*/
DOTDetector::DOTTemplate::DOTTemplate() : texturelessRatio(-1.f) {}
DOTDetector::DOTTemplate::DOTTemplate( const cv::Mat& _quantizedImage, int _classID, const cv::Mat& _maskedImage, const cv::Mat& _gradientMask ) :
quantizedImage(_quantizedImage), texturelessRatio(computeTexturelessRatio(_quantizedImage))
DOTDetector::DOTTemplate::TrainData::TrainData()
{
}
DOTDetector::DOTTemplate::TrainData::TrainData( const Mat& _maskedImage, const cv::Mat& _strongestGradientsMask )
: maskedImage( _maskedImage ), strongestGradientsMask( _strongestGradientsMask )
{
addClassID( _classID, _maskedImage, _gradientMask );
}
void DOTDetector::DOTTemplate::addClassID( int _classID, const cv::Mat& _maskedImage, const cv::Mat& _gradientMask )
DOTDetector::DOTTemplate::DOTTemplate() : texturelessRatio(-1.f) {}
DOTDetector::DOTTemplate::DOTTemplate( const cv::Mat& _quantizedImage, int _objectClassID, const cv::Mat& _maskedImage, const cv::Mat& _strongestGradientsMask ) :
quantizedImage(_quantizedImage), texturelessRatio(computeTexturelessRatio(_quantizedImage))
{
CV_Assert( _classID >= 0 );
addObjectClassID( _objectClassID, _maskedImage, _strongestGradientsMask );
}
void DOTDetector::DOTTemplate::addObjectClassID( int _objectClassID, const cv::Mat& _maskedImage, const cv::Mat& _strongestGradientsMask )
{
CV_Assert( _objectClassID >= 0 );
bool isFound = false;
for( size_t i = 0; i < classIDs.size(); i++ )
for( size_t i = 0; i < objectClassIDs.size(); i++ )
{
if( classIDs[i] == _classID )
if( objectClassIDs[i] == _objectClassID )
{
isFound = true;
break;
@@ -508,16 +517,28 @@ void DOTDetector::DOTTemplate::addClassID( int _classID, const cv::Mat& _maskedI
if( !isFound )
{
classIDs.push_back( _classID );
objectClassIDs.push_back( _objectClassID );
if( !_maskedImage.empty() )
{
CV_Assert( !_gradientMask.empty() );
maskedImages.push_back( _maskedImage );
gradientMasks.push_back( _gradientMask );
CV_Assert( !_strongestGradientsMask.empty() );
trainData.push_back( TrainData(_maskedImage, _strongestGradientsMask) );
}
}
}
const DOTDetector::DOTTemplate::TrainData* DOTDetector::DOTTemplate::getTrainData( int objectClassID ) const
{
if( objectClassID >= 0 )
{
for( size_t i = 0; i < objectClassIDs.size(); i++ )
{
if( objectClassID == objectClassIDs[i] )
return &trainData[i];
}
}
return 0;
}
float DOTDetector::DOTTemplate::computeTexturelessRatio( const cv::Mat& quantizedImage )
{
const uchar TEXTURELESS_VAL = 1 << DOTDetector::TrainParams::BIN_COUNT;
@@ -536,14 +557,14 @@ float DOTDetector::DOTTemplate::computeTexturelessRatio( const cv::Mat& quantize
void DOTDetector::DOTTemplate::read( FileNode& fn )
{
fn["template"] >> quantizedImage;
fn["classIDs"] >> classIDs;
fn["objectClassIDs"] >> objectClassIDs;
texturelessRatio = fn["texturelessRatio"];
}
void DOTDetector::DOTTemplate::write( FileStorage& fs ) const
{
fs << "template" << quantizedImage;
fs << "classIDs" << classIDs;
fs << "objectClassIDs" << objectClassIDs;
fs << "texturelessRatio" << texturelessRatio;
}
@@ -551,11 +572,11 @@ void DOTDetector::DOTTemplate::write( FileStorage& fs ) const
* DOTDetector
*/
DOTDetector::DOTDetector() : isAddImageAndGradientMask( false )
DOTDetector::DOTDetector()
{
}
DOTDetector::DOTDetector( const std::string& filename ) : isAddImageAndGradientMask( false )
DOTDetector::DOTDetector( const std::string& filename )
{
load( filename );
}
@@ -567,7 +588,7 @@ DOTDetector::~DOTDetector()
void DOTDetector::clear()
{
classNames.clear();
objectClassNames.clear();
dotTemplates.clear();
}
@@ -586,7 +607,7 @@ void DOTDetector::read( FileNode& fn )
{
string name;
fni >> name;
classNames.push_back( name );
objectClassNames.push_back( name );
}
// read DOT templates
@@ -608,11 +629,11 @@ void DOTDetector::write( FileStorage& fs ) const
fs << "}"; //params
// write class names
fs << "class_count" << (int)classNames.size();
fs << "class_count" << (int)objectClassNames.size();
fs << "class_names" << "[";
for( size_t i = 0; i < classNames.size(); i++ )
for( size_t i = 0; i < objectClassNames.size(); i++ )
{
fs << classNames[i];
fs << objectClassNames[i];
}
fs << "]";
@@ -647,29 +668,31 @@ void DOTDetector::save( const std::string& filename ) const
}
}
void DOTDetector::train( const string& _baseDirName, const TrainParams& _trainParams, bool /*_isAddImageAndGradientMask*/ )
void DOTDetector::train( const string& _baseDirName, const TrainParams& _trainParams, bool isAddImageAndGradientMask )
{
clear();
trainParams = _trainParams;
trainParams.asserts();
trainParams.isConsistent();
string baseDirName = _baseDirName + (*(_baseDirName.end()-1) == '/' ? "" : "/");
const int regionSize_2 = trainParams.regionSize / 2;
readDirContent( baseDirName+"objects.txt", classNames );
vector<string> allObjectClassNames;
readDirContent( baseDirName + "objects.txt", allObjectClassNames );
for( size_t objIdx = 0; objIdx < classNames.size(); objIdx++ )
for( size_t objIdx = 0; objIdx < allObjectClassNames.size(); objIdx++ )
{
string curObjDirName = baseDirName + classNames[objIdx] + "/";
string curObjDirName = baseDirName + allObjectClassNames[objIdx] + "/";
cout << "===============" << classNames[objIdx] << "===============" << endl;
cout << "===============" << allObjectClassNames[objIdx] << "===============" << endl;
vector<string> imagesFilenames;
readDirContent( curObjDirName + "images.txt", imagesFilenames );
if( imagesFilenames.empty() )
continue;
objectClassNames.push_back( allObjectClassNames[objIdx] );
int countSamples = 0;
for( size_t imgIdx = 0; imgIdx < imagesFilenames.size(); imgIdx++ )
{
@@ -692,47 +715,62 @@ void DOTDetector::train( const string& _baseDirName, const TrainParams& _trainPa
countSamples++;
Mat trainImageExt, trainMaskExt, trainQuantizedImage, detectQuantizedImage;
Mat trainImageExt, trainMaskExt, trainQuantizedImage, queryQuantizedImage;
Mat trainMagnitudesExt, trainAnglesExt;
computeWinData( image, mask, trainParams.winSize,
trainImageExt, trainMaskExt,
trainMagnitudesExt, trainAnglesExt, regionSize_2 );
static int index_ = 0;
{
stringstream ss;
ss << "/files/Datasets/test_temp/" << index_ << ".png";
index_++;
imwrite( ss.str(), trainImageExt );
}
quantizeToTrain( trainMagnitudesExt, trainAnglesExt, trainMaskExt, trainQuantizedImage, trainParams );
quantizeToDetect( trainMagnitudesExt, trainAnglesExt, detectQuantizedImage, trainParams );
quantizeToDetect( trainMagnitudesExt, trainAnglesExt, queryQuantizedImage,
trainParams.regionSize, trainParams );
vector<vector<Rect> > rects;
vector<vector<float> > ratios;
vector<vector<int> > trainTemplatesIdxs;
detectQuantized( detectQuantizedImage, trainParams.minRatio, rects, ratios, trainTemplatesIdxs );
vector<vector<int> > dotTemplateIndices;
Mat maskedTrainImage, trainGradientMask;
detectQuantized( queryQuantizedImage, trainParams.minRatio, rects, ratios, dotTemplateIndices );
Mat trainMaskedImage, trainStrongestGradientMask;
if( isAddImageAndGradientMask )
{
trainImageExt.copyTo( maskedTrainImage, trainMaskExt);
trainImageExt.copyTo( trainMaskedImage, trainMaskExt );
computeTrainUsedStrongestMask( trainMagnitudesExt, trainAnglesExt, trainMaskExt, trainQuantizedImage,
trainGradientMask, trainParams.regionSize, trainParams.minMagnitude );
trainStrongestGradientMask, trainParams.regionSize, trainParams.minMagnitude );
}
int classID = classNames.size()-1;
int objectClassID = objectClassNames.size()-1;
bool isFound = false;
for( size_t cIdx = 0; cIdx < trainTemplatesIdxs.size(); cIdx++ )
for( size_t cIdx = 0; cIdx < dotTemplateIndices.size(); cIdx++ )
{
if( trainTemplatesIdxs[cIdx].size() )
if( dotTemplateIndices[cIdx].size() )
{
for( size_t i = 0; i < trainTemplatesIdxs[cIdx].size(); i++ )
for( size_t i = 0; i < dotTemplateIndices[cIdx].size(); i++ )
{
int tIdx = trainTemplatesIdxs[cIdx][i];
int tIdx = dotTemplateIndices[cIdx][i];
dotTemplates[tIdx].addClassID( classID, maskedTrainImage, trainGradientMask );
if( isAddImageAndGradientMask )
dotTemplates[tIdx].addObjectClassID( objectClassID, trainMaskedImage, trainStrongestGradientMask );
else
dotTemplates[tIdx].addObjectClassID( objectClassID );
isFound = true;
}
}
}
if( !isFound )
{
dotTemplates.push_back( DOTTemplate(trainQuantizedImage, classID, maskedTrainImage, trainGradientMask) );
if( isAddImageAndGradientMask )
dotTemplates.push_back( DOTTemplate(trainQuantizedImage, objectClassID, trainMaskedImage, trainStrongestGradientMask) );
else
dotTemplates.push_back( DOTTemplate(trainQuantizedImage, objectClassID ) );
}
cout << "dot templates size = " << dotTemplates.size() << endl;
@@ -740,8 +778,10 @@ void DOTDetector::train( const string& _baseDirName, const TrainParams& _trainPa
}
}
void DOTDetector::detectQuantized( const Mat& testQuantizedImage, float minRatio,
vector<vector<Rect> >& rects, vector<vector<float> >& ratios, vector<vector<int> >& trainTemlateIdxs ) const
void DOTDetector::detectQuantized( const Mat& queryQuantizedImage, float minRatio,
vector<vector<Rect> >& rects,
vector<vector<float> >& ratios,
vector<vector<int> >& dotTemplateIndices ) const
{
if( dotTemplates.empty() )
return;
@@ -749,29 +789,29 @@ void DOTDetector::detectQuantized( const Mat& testQuantizedImage, float minRatio
const int regionsPerRow = dotTemplates[0].quantizedImage.rows;
const int regionsPerCol = dotTemplates[0].quantizedImage.cols;
int classCount = classNames.size();
int objectClassCount = objectClassNames.size();
rects.resize( classCount );
ratios.resize( classCount );
trainTemlateIdxs.resize( classCount );
rects.resize( objectClassCount );
ratios.resize( objectClassCount );
dotTemplateIndices.resize( objectClassCount );
for( size_t tIdx = 0; tIdx < dotTemplates.size(); tIdx++ )
{
Rect r( 0, 0, regionsPerCol, regionsPerRow );
for( r.y = 0; r.y <= testQuantizedImage.rows-r.height; r.y++ )
for( r.y = 0; r.y <= queryQuantizedImage.rows-r.height; r.y++ )
{
for( r.x = 0; r.x <= testQuantizedImage.cols-r.width; r.x++ )
for( r.x = 0; r.x <= queryQuantizedImage.cols-r.width; r.x++ )
{
float ratio, texturelessRatio;
andQuantizedImages( testQuantizedImage(r), dotTemplates[tIdx].quantizedImage, ratio, texturelessRatio );
andQuantizedImages( queryQuantizedImage(r), dotTemplates[tIdx].quantizedImage, ratio, texturelessRatio );
if( ratio > minRatio && texturelessRatio < dotTemplates[tIdx].texturelessRatio )
{
for( size_t cIdx = 0; cIdx < dotTemplates[tIdx].classIDs.size(); cIdx++ )
for( size_t cIdx = 0; cIdx < dotTemplates[tIdx].objectClassIDs.size(); cIdx++ )
{
int classID = dotTemplates[tIdx].classIDs[cIdx];
rects[classID].push_back( r );
ratios[classID].push_back( ratio );
trainTemlateIdxs[classID].push_back( tIdx );
int objectClassID = dotTemplates[tIdx].objectClassIDs[cIdx];
rects[objectClassID].push_back( r );
ratios[objectClassID].push_back( ratio );
dotTemplateIndices[objectClassID].push_back( tIdx );
}
}
}
@@ -780,23 +820,23 @@ void DOTDetector::detectQuantized( const Mat& testQuantizedImage, float minRatio
}
void DOTDetector::detectMultiScale( const Mat& image, vector<vector<Rect> >& rects,
const DetectParams& detectParams, vector<vector<float> >* ratios, vector<vector<int> >* trainTemplateIndices ) const
const DetectParams& detectParams, vector<vector<float> >* ratios, vector<vector<int> >* dotTemplateIndices ) const
{
detectParams.asserts( trainParams.minRatio );
detectParams.isConsistent( trainParams.minRatio );
int classCount = classNames.size();
rects.resize( classCount );
int objectClassCount = objectClassNames.size();
rects.resize( objectClassCount );
if( ratios )
{
ratios->clear();
if( !detectParams.isGroup )
ratios->resize( classCount );
ratios->resize( objectClassCount );
}
if( trainTemplateIndices )
if( dotTemplateIndices )
{
trainTemplateIndices->clear();
dotTemplateIndices->clear();
if( !detectParams.isGroup )
trainTemplateIndices->resize( classCount );
dotTemplateIndices->resize( objectClassCount );
}
Mat magnitudes, angles;
@@ -806,11 +846,11 @@ void DOTDetector::detectMultiScale( const Mat& image, vector<vector<Rect> >& rec
Mat quantizedImage;
vector<vector<Rect> > curRects;
vector<vector<float> > curRatios;
vector<vector<int> > curTrainTemlateIdxs;
quantizeToDetect( magnitudes, angles, quantizedImage, trainParams );
detectQuantized( quantizedImage, detectParams.minRatio, curRects, curRatios, curTrainTemlateIdxs );
vector<vector<int> > curDotTemlateIndices;
quantizeToDetect( magnitudes, angles, quantizedImage, regionSize, trainParams );
detectQuantized( quantizedImage, detectParams.minRatio, curRects, curRatios, curDotTemlateIndices );
for( int ci = 0; ci < classCount; ci++ )
for( int ci = 0; ci < objectClassCount; ci++ )
{
for( size_t ri = 0; ri < curRects[ci].size(); ri++ )
{
@@ -823,8 +863,8 @@ void DOTDetector::detectMultiScale( const Mat& image, vector<vector<Rect> >& rec
rects[ci].push_back( r );
if( ratios && !detectParams.isGroup )
(*ratios)[ci].push_back( curRatios[ci][ri] );
if( trainTemplateIndices && !detectParams.isGroup )
(*trainTemplateIndices)[ci].push_back( curTrainTemlateIdxs[ci][ri] );
if( dotTemplateIndices && !detectParams.isGroup )
(*dotTemplateIndices)[ci].push_back( curDotTemlateIndices[ci][ri] );
}
}
}
@@ -839,9 +879,9 @@ const vector<DOTDetector::DOTTemplate>& DOTDetector::getDOTTemplates() const
return dotTemplates;
}
const vector<string>& DOTDetector::getClassNames() const
const vector<string>& DOTDetector::getObjectClassNames() const
{
return classNames;
return objectClassNames;
}
void DOTDetector::groupRectanglesList( std::vector<std::vector<cv::Rect> >& rectList, int groupThreshold, double eps )