Revision history [back]

you can do the following 1-find all Regions in the image by using findcontour 2-find center for every region, you can use moments , the center point will represent the region 3- Pass the centers to Kmeans Kmean clustring or this link
Kmean will find the clusters group of points 4-For every cluster You can use one of the following function to detect the rectangle (boundingRect,approxPolyDP,RotatedRect)

you can do the following following

1. 1-find all Regions in the image by by using findcontour 2-find center for every region, you can use use moments , the center point will represent represent the region 3- Pass the centers to Kmeans Kmean Kmean clustring or this link
   

   Kmean will find the clusters  group of points 
    4-For every cluster You can use one of the following function to to
   

   detect the rectangle rectangle (boundingRect,approxPolyDP,RotatedRect)

you can do the following

1. 1-find find all Regions in the image by using findcontour 2-find using findcontour

2. find center for every region, you you can use moments , the center point will represent the region 3- region

3. Pass the centers to Kmeans Kmean Kmean
   clustring or this link or this link

   Kmean will find the clusters  group of points 
   

4-For

4. For every cluster You can use one of the following function to

   detect todetect the rectangle (boundingRect,approxPolyDP,RotatedRect)rectangle(boundingRect, approxPolyDP,rotatedRect)