A genetic algorithm approach to edge detection on image processing applications

Computer vision is currently used in the industrial environment to inspect the quality of manufactured pieces. Such applications are highly dependent on image processing capabilities, particularly, edge detection. This paper introduces a method based on genetic algorithms to identify, edges in a picture. The genetic algorithm randomly establishes a set of potential edge points and through all iterative process determines those with better probability of belonging to the actual edges in the image. A linear time algorithm is then used to extrapolate the remaining points. A detailed description of the method is presented with a practical example.