Add method for simple distance

async_detector/AsyncDetector.cpp

@@ -298,7 +298,7 @@ void AsyncDetector::CaptureThread(std::string fileName, int startFrame, float* f
     const int minStaticTime = 5;
 
     TrackerSettings trackerSettings;
-	trackerSettings.SetDistances({ 1.0f, 0.0f, 0.0f, 0.0f, 0.0f });
+	trackerSettings.SetDistance(tracking::DistCenters);
     trackerSettings.m_kalmanType = tracking::KalmanLinear;
     trackerSettings.m_filterGoal = tracking::FilterRect;
     trackerSettings.m_lostTrackType = tracking::TrackKCF; // Use KCF tracker for collisions resolving

cars_counting/CarsCounting.cpp

@@ -259,7 +259,7 @@ bool CarsCounting::InitTracker(cv::UMat frame)
     m_detector->SetMinObjectSize(cv::Size(m_minObjWidth, m_minObjWidth));
 
     TrackerSettings settings;
-	settings.SetDistances({ 1.0f, 0.0f, 0.0f, 0.0f, 0.0f });
+	settings.SetDistance(tracking::DistCenters);
     settings.m_kalmanType = tracking::KalmanLinear;
     settings.m_filterGoal = tracking::FilterRect;
     settings.m_lostTrackType = tracking::TrackCSRT; // Use KCF tracker for collisions resolving

example/MouseExample.h

@@ -47,7 +47,7 @@ void MouseTracking(cv::CommandLineParser parser)
 	cv::setMouseCallback("Video", mv_MouseCallback, (void*)&pointXY);
 
 	TrackerSettings settings;
-	settings.SetDistances({ 1.0f, 0.0f, 0.0f, 0.0f, 0.0f });
+	settings.SetDistance(tracking::DistCenters);
 	settings.m_kalmanType = tracking::KalmanLinear;
 	settings.m_filterGoal = tracking::FilterCenter;
 	settings.m_lostTrackType = tracking::TrackNone;

example/examples.h

@@ -60,14 +60,14 @@ class MotionDetectorExample : public VideoExample
     bool InitTracker(cv::UMat frame)
     {
         TrackerSettings settings;
-		settings.SetDistances({ 0.0f, 0.0f, 1.0f, 0.0f, 0.0f });
+		settings.SetDistance(tracking::DistCenters);
         settings.m_kalmanType = tracking::KalmanLinear;
         settings.m_filterGoal = tracking::FilterRect;
         settings.m_lostTrackType = tracking::TrackKCF;       // Use visual objects tracker for collisions resolving
         settings.m_matchType = tracking::MatchHungrian;
         settings.m_dt = 0.4f;                             // Delta time for Kalman filter
         settings.m_accelNoiseMag = 0.5f;                  // Accel noise magnitude for Kalman filter
-		settings.m_distThres = 0.6f; // frame.rows / 10.f;         // Distance threshold between region and object on two frames
+		settings.m_distThres = frame.rows / 10.f;         // Distance threshold between region and object on two frames
 
         settings.m_useAbandonedDetection = false;
         if (settings.m_useAbandonedDetection)
@@ -175,7 +175,7 @@ class FaceDetectorExample : public VideoExample
     bool InitTracker(cv::UMat frame)
     {
         TrackerSettings settings;
-		settings.SetDistances({ 0.0f, 0.0f, 1.0f, 0.0f, 0.0f });
+		settings.SetDistance(tracking::DistJaccard);
         settings.m_kalmanType = tracking::KalmanUnscented;
         settings.m_filterGoal = tracking::FilterRect;
         settings.m_lostTrackType = tracking::TrackCSRT;       // Use visual objects tracker for collisions resolving
@@ -271,7 +271,7 @@ class PedestrianDetectorExample : public VideoExample
     bool InitTracker(cv::UMat frame)
     {
         TrackerSettings settings;
-		settings.SetDistances({ 0.0f, 1.0f, 0.0f, 0.0f, 0.0f });
+		settings.SetDistance(tracking::DistRects);
         settings.m_kalmanType = tracking::KalmanLinear;
         settings.m_filterGoal = tracking::FilterRect;
         settings.m_lostTrackType = tracking::TrackCSRT;       // Use visual objects tracker for collisions resolving
@@ -368,7 +368,7 @@ class SSDMobileNetExample : public VideoExample
     bool InitTracker(cv::UMat frame)
     {
         TrackerSettings settings;
-		settings.SetDistances({ 0.0f, 1.0f, 0.0f, 0.0f, 0.0f });
+		settings.SetDistance(tracking::DistRects);
         settings.m_kalmanType = tracking::KalmanLinear;
         settings.m_filterGoal = tracking::FilterRect;
         settings.m_lostTrackType = tracking::TrackCSRT;       // Use visual objects tracker for collisions resolving
@@ -492,7 +492,7 @@ class YoloExample : public VideoExample
     bool InitTracker(cv::UMat frame)
     {
         TrackerSettings settings;
-		settings.SetDistances({ 0.0f, 1.0f, 0.0f, 0.0f, 0.0f });
+		settings.SetDistance(tracking::DistRects);
         settings.m_kalmanType = tracking::KalmanLinear;
         settings.m_filterGoal = tracking::FilterRect;
         settings.m_lostTrackType = tracking::TrackCSRT;       // Use visual objects tracker for collisions resolving
@@ -616,7 +616,7 @@ class YoloDarknetExample : public VideoExample
     bool InitTracker(cv::UMat frame)
 	{
 		TrackerSettings settings;
-		settings.SetDistances({ 0.0f, 1.0f, 0.0f, 0.0f, 0.0f });
+		settings.SetDistance(tracking::DistRects);
 		settings.m_kalmanType = tracking::KalmanLinear;
 		settings.m_filterGoal = tracking::FilterRect;
         settings.m_lostTrackType = tracking::TrackKCF;       // Use visual objects tracker for collisions resolving

src/Tracker/Ctracker.cpp

@@ -123,9 +123,9 @@ void CTracker::UpdateTrackingState(
     }
 
     // Update Kalman Filters state
-    const ptrdiff_t stop_i = static_cast<int>(assignment.size());
+    const ptrdiff_t stop_i = static_cast<ptrdiff_t>(assignment.size());
 #pragma omp parallel for
-    for (int i = 0; i < stop_i; ++i)
+    for (ptrdiff_t i = 0; i < stop_i; ++i)
     {
         // If track updated less than one time, than filter state is not correct.
         if (assignment[i] != -1) // If we have assigned detect, then update using its coordinates,
@@ -156,11 +156,11 @@ void CTracker::CreateDistaceMatrix(const regions_t& regions, distMatrix_t& costM
     const size_t N = m_tracks.size();	// Tracking objects
     maxCost = 0;
 
-	for (size_t i = 0; i < m_tracks.size(); i++)
+	for (size_t i = 0; i < m_tracks.size(); ++i)
 	{
 		const auto& track = m_tracks[i];
 
-		for (size_t j = 0; j < regions.size(); j++)
+		for (size_t j = 0; j < regions.size(); ++j)
 		{
 			auto dist = maxPossibleCost;
 			if (m_tracks[i]->CheckType(regions[j].m_type))

src/Tracker/Ctracker.h

@@ -88,8 +88,8 @@ struct TrackerSettings
 	bool CheckDistance() const
 	{
 		track_t sum = std::accumulate(m_distType.begin(), m_distType.end(), 0.0f);
-		track_t maxOne = std::max(1.0f, std::fabs(sum));
-		return std::fabs(sum - 1.0f) <= std::numeric_limits<track_t>::epsilon() * maxOne;
+		track_t maxOne = std::max(1.0f, std::fabs(sum));
+		return std::fabs(sum - 1.0f) <= std::numeric_limits<track_t>::epsilon() * maxOne;
 	}
 
 	///
@@ -105,6 +105,14 @@ struct TrackerSettings
 		}
 		return res;
 	}
+
+	///
+	bool SetDistance(tracking::DistType distType)
+	{
+		std::fill(m_distType.begin(), m_distType.end(), 0.0f);
+		m_distType[distType] = 1.f;
+		return true;
+	}
 };
 
 ///
@@ -114,9 +122,9 @@ class CTracker
 {
 public:
     CTracker(const TrackerSettings& settings);
-	CTracker(const CTracker&) = delete;
-	CTracker(CTracker&&) = delete;
-	CTracker& operator=(const CTracker&) = delete;
+	CTracker(const CTracker&) = delete;
+	CTracker(CTracker&&) = delete;
+	CTracker& operator=(const CTracker&) = delete;
 	CTracker& operator=(CTracker&&) = delete;
 
 	~CTracker(void);

src/Tracker/track.cpp

@@ -101,28 +101,29 @@ track_t CTrack::CalcDistHist(const CRegion& reg, cv::UMat currFrame) const
 {
 	track_t res = 1;
 
-	if (reg.m_hist.empty())
-	{
-		int bins = 64;
-		std::vector<int> histSize;
-		std::vector<float> ranges;
-		std::vector<int> channels;
-
-		for (int i = 0, stop = currFrame.channels(); i < stop; ++i)
-		{
-			histSize.push_back(bins);
-			ranges.push_back(0);
-			ranges.push_back(255);
-			channels.push_back(i);
-		}
-
-		std::vector<cv::UMat> regROI = { currFrame(reg.m_brect) };
-		cv::calcHist(regROI, channels, cv::Mat(), reg.m_hist, histSize, ranges, false);
+	if (reg.m_hist.empty())
+	{
+		int bins = 64;
+		std::vector<int> histSize;
+		std::vector<float> ranges;
+		std::vector<int> channels;
+
+		for (int i = 0, stop = currFrame.channels(); i < stop; ++i)
+		{
+			histSize.push_back(bins);
+			ranges.push_back(0);
+			ranges.push_back(255);
+			channels.push_back(i);
+		}
+
+		std::vector<cv::UMat> regROI = { currFrame(reg.m_brect) };
+		cv::calcHist(regROI, channels, cv::Mat(), reg.m_hist, histSize, ranges, false);
 		cv::normalize(reg.m_hist, reg.m_hist, 0, 1, cv::NORM_MINMAX, -1, cv::Mat());
 	}
 	if (!reg.m_hist.empty() && !m_lastRegion.m_hist.empty())
 	{
 		res = static_cast<track_t>(cv::compareHist(reg.m_hist, m_lastRegion.m_hist, CV_COMP_BHATTACHARYYA));
+		//res = 1.f - static_cast<track_t>(cv::compareHist(reg.m_hist, m_lastRegion.m_hist, CV_COMP_CORREL));
 	}
 
 	return res;