Cellular edge detection: Combining cellular automata and cellular learning automata

Abstract

In this paper, we propose a cellular edge detection (CED) algorithm which utilizes cellular automata (CA) and cellular learning automata (CLA). The CED algorithm is an adaptive, intelligent and learnable algorithm for edge detection of binary and grayscale images. Here, we introduce a new CA local rule with adaptive neighborhood type to produce the edge map of image as opposed to CA with fixed neighborhood. The proposed adaptive algorithm uses the von Neumann and Moore neighborhood types. Experimental results demonstrate that the CED algorithm has superior accuracy and performance in contrast to other edge detection methods such as Sobel, Prewitt, Robert, LoG and Canny operators. Moreover, the CED algorithm loses fewer details while extracting image edges compare to other edge detection methods.