During the time I spent today on my OSS, I read the “Comparative Study of Hough Transform Methods for Circle Finding”. In short it is an overview of Hough Transforms used to find circles within images. The reason I started reading it is that it was referenced in the OpenCV documentation as the root of their algorithm for the HoughCircles function. After delving into the paper abit I realized that I need to order two more articles to actually understand the algorithm being used.
After spending some time reading the paper, I again turned my attention to working on my target finding algorithm. I am having a bit of difficulty of how I should establish my target from other circles included in the scene. I would really like to get back all of the circles in the image but the algorithm seems to do a poor job returning concentric circles. If I could get the concentric circles, I would compare their radii to determine my target from others. I think I need to lookup point clustering algorithms.