Method for Edges Detection in Digital Images Through the Use of Cellular Automata

Abstract

In the present article, an algorithm based on cellular automata for noise elimination and edge detection in grayscale images is proposed. However, the focus of this project will be on the process of identifying contours, since this represents a higher challenge at the research level. Also, the cellular automaton has an adaptive behavior, which allows it to expand when it considers that the information coming from its initial neighbors is insufficient to determine if the pixel in evaluation is a “border pixel” or not, as this is an important feature together with a set of transition rules useful to accentuate relevant details within the image. By integrating these characteristics, we find the results obtained by the Proposed Algorithm presented greater similarity compared to images with ideal borders, since the edge recovery ranges between 92.30% and 97.21%, which indicates that MEDCA is in general terms, more efficient compared to similar algorithms.