Study of Edge Detection Algorithms

Abstract: Edge detection is a very important area in the field of Computer Vision. Edges define the boundEdge detectors are local image processing methods designed to detect edge pixels in an image. Edge detection is the approach used most frequently for segmenting images based on abrupt changes in intensity. The best edge detector is the human eye.An edge represents the boundaries of an object in an image, there by representing the structural properties of the object. This paper shows the comparisons among the various edge detection algorithms. From the results it is observed that the canny edge detector performance is far superior than the Sobel, Prewitt, Roberts and Laplacian of Gaussian. (LoG). Keywords: Edge Detection, Sobel, Prewitt, Roberts, LoG, Canny.