Reverted back to the previous method where a public method groupRectangles was added to HOGDescriptor to take care of ROI and weight grouping

modules/objdetect/include/opencv2/objdetect/objdetect.hpp

@@ -327,6 +327,23 @@ private:
     vector<string> classNames;
 };
 
+// class for grouping object candidates, detected by Cascade Classifier, HOG etc.
+// instance of the class is to be passed to cv::partition (see cxoperations.hpp)
+class CV_EXPORTS SimilarRects
+{
+public:
+    SimilarRects(double _eps) : eps(_eps) {}
+    inline bool operator()(const Rect& r1, const Rect& r2) const
+    {
+        double delta = eps*(std::min(r1.width, r2.width) + std::min(r1.height, r2.height))*0.5;
+        return std::abs(r1.x - r2.x) <= delta &&
+            std::abs(r1.y - r2.y) <= delta &&
+            std::abs(r1.x + r1.width - r2.x - r2.width) <= delta &&
+            std::abs(r1.y + r1.height - r2.y - r2.height) <= delta;
+    }
+    double eps;
+};
+
 CV_EXPORTS void groupRectangles(CV_OUT CV_IN_OUT vector<Rect>& rectList, int groupThreshold, double eps=0.2);
 CV_EXPORTS_W void groupRectangles(CV_OUT CV_IN_OUT vector<Rect>& rectList, CV_OUT vector<int>& weights, int groupThreshold, double eps=0.2);
 CV_EXPORTS void groupRectangles( vector<Rect>& rectList, int groupThreshold, double eps, vector<int>* weights, vector<double>* levelWeights );
@@ -611,6 +628,7 @@ public:
 
    // read/parse Dalal's alt model file
    void readALTModel(std::string modelfile);
+   void groupRectangles(vector<cv::Rect>& rectList, vector<double>& weights, int groupThreshold, double eps) const;
 };
 
 

modules/objdetect/src/cascadedetect.cpp

@@ -114,24 +114,6 @@ struct Logger
 namespace cv
 {
 
-// class for grouping object candidates, detected by Cascade Classifier, HOG etc.
-// instance of the class is to be passed to cv::partition (see cxoperations.hpp)
-class CV_EXPORTS SimilarRects
-{
-public:
-    SimilarRects(double _eps) : eps(_eps) {}
-    inline bool operator()(const Rect& r1, const Rect& r2) const
-    {
-        double delta = eps*(std::min(r1.width, r2.width) + std::min(r1.height, r2.height))*0.5;
-        return std::abs(r1.x - r2.x) <= delta &&
-        std::abs(r1.y - r2.y) <= delta &&
-        std::abs(r1.x + r1.width - r2.x - r2.width) <= delta &&
-        std::abs(r1.y + r1.height - r2.y - r2.height) <= delta;
-    }
-    double eps;
-};
-
-
 void groupRectangles(vector<Rect>& rectList, int groupThreshold, double eps, vector<int>* weights, vector<double>* levelWeights)
 {
     if( groupThreshold <= 0 || rectList.empty() )

modules/objdetect/src/hog.cpp

@@ -1060,9 +1060,7 @@ void HOGDescriptor::detectMultiScale(
     }
     else
     {
-        vector<int> dummy;
-        dummy.resize(foundLocations.size(), INT_MAX);
-        groupRectangles(foundLocations, (int)finalThreshold, 0.2, &dummy, &foundWeights);
+        groupRectangles(foundLocations, foundWeights, (int)finalThreshold, 0.2);
     }
 }
 
@@ -2636,4 +2634,82 @@ void HOGDescriptor::readALTModel(std::string modelfile)
    fclose(modelfl);
 }
 
+void HOGDescriptor::groupRectangles(vector<cv::Rect>& rectList, vector<double>& weights, int groupThreshold, double eps) const
+{
+    if( groupThreshold <= 0 || rectList.empty() )
+    {
+        return;
+    }
+
+    CV_Assert(rectList.size() == weights.size());
+
+    vector<int> labels;
+    int nclasses = partition(rectList, labels, SimilarRects(eps));
+
+    vector<cv::Rect_<double>> rrects(nclasses);
+    vector<int> numInClass(nclasses, 0);
+    vector<double> foundWeights(nclasses, DBL_MIN);
+    vector<double> totalFactorsPerClass(nclasses, 1);
+    int i, j, nlabels = (int)labels.size();
+
+    for( i = 0; i < nlabels; i++ )
+    {
+        int cls = labels[i];
+        rrects[cls].x += rectList[i].x;
+        rrects[cls].y += rectList[i].y;
+        rrects[cls].width += rectList[i].width;
+        rrects[cls].height += rectList[i].height;
+        foundWeights[cls] = max(foundWeights[cls], weights[i]);
+        numInClass[cls]++;
+    }
+
+    for( i = 0; i < nclasses; i++ )
+    {
+        // find the average of all ROI in the cluster
+        cv::Rect_<double> r = rrects[i];
+        float s = 1.f/numInClass[i];
+        rrects[i] = cv::Rect_<double>(cv::saturate_cast<double>(r.x*s),
+            cv::saturate_cast<double>(r.y*s),
+            cv::saturate_cast<double>(r.width*s),
+            cv::saturate_cast<double>(r.height*s));
+    }
+
+    rectList.clear();
+    weights.clear();
+
+    for( i = 0; i < nclasses; i++ )
+    {
+        cv::Rect r1 = rrects[i];
+        int n1 = numInClass[i];
+        double w1 = foundWeights[i];
+        if( n1 <= groupThreshold )
+            continue;
+        // filter out small rectangles inside large rectangles
+        for( j = 0; j < nclasses; j++ )
+        {
+            int n2 = numInClass[j];
+
+            if( j == i || n2 <= groupThreshold )
+                continue;
+
+            cv::Rect r2 = rrects[j];
+
+            int dx = cv::saturate_cast<int>( r2.width * eps );
+            int dy = cv::saturate_cast<int>( r2.height * eps );
+
+            if( r1.x >= r2.x - dx &&
+                r1.y >= r2.y - dy &&
+                r1.x + r1.width <= r2.x + r2.width + dx &&
+                r1.y + r1.height <= r2.y + r2.height + dy &&
+                (n2 > std::max(3, n1) || n1 < 3) )
+                break;
+        }
+
+        if( j == nclasses )
+        {
+            rectList.push_back(r1);
+            weights.push_back(w1);
+        }
+    }
+}
 }