Extension of the Canny algorithm for the efficient color edge detection

The edge detection is a key component for many modern computer vision applications while it is certainly not the only way to identify an object, or track a feature. It can be one of the most convenient methods to recognize objects in an image if it is done quickly and consistently. Many algorithms proposed are applied to gray level images. But, there are limits in method using only intensity information. So, many researchers have tried to do researches using color information. In this paper, we propose the new color edge detection method which is the extended Canny algorithm. Compared with the existing algorithms in performance, the proposed algorithm is better than others in fine details, shaded region, and image with almost same intensity values.