A Review Paper: On Various Edge Detection Techniques

Abstract

Edge detection is an important phenomenon in image segmentation. In general kernel based methods like Sobel, Canny, Roberts etc. are used which are based on first and second derivatives pixels intensity. Due to the some of the limitations of such approaches soft computing based methods are also evolved. This paper presented an overview of the kernel and soft computing based methods along with notable research done in the field of edge detection. Basis analogies of the methods have been described along with literature survey of each class of methods.