A hybrid method realizing shock filter by Cellular Neural Network

Abstract

Shock filter is an important type of the filters in image processing. In recent years, research on shock filters has been an interested topic in signal and image processing. In this paper we propose a method to realize the shock filter by CNN (Cellular Neural Network). The method offers fast processing time in comparison with that on the serial digital computers and can be used in many real-time image processing systems. We focus on development of the shock filter algorithm realizing edge enhancement and de-noising on the Bi-I CNN visual computer. Besides presenting the algorithm, many experiments on Bi-I are given for demonstration.