Detecting ellipses of limited eccentricity in images with high noise levels

摘要

We present a newly developed algorithm for the detection of elliptical shapes in images in the presence of high noise levels. The algorithm combines a modified version of the Hough transform with a genetic algorithm, namely Differential Evolution. Suggestions for a parallel implementation are given. In our implementation the algorithm is restricted due to the technical problem to be solved, yet it can be easily generalized to arbitrary ellipse detection.