Abstract
Circle detection is important for many applications. We propose a method for circle detection using a histogram based accumulator space. Our method uses a single accumulator space to find different size circles. It is tested to be robust and able to detect partial circles or circles in partial occlusion. The method is based on finding the distance transform between pixels and edge pixels in the image. A histogram is then used to count the frequency of the distances that participates in the accumulator space votes.