A bio-inspired neural model for colour image segmentation

Abstract

This paper describes a multi-scale neural model to enhance regions and extract contours of colour-texture image taking into consideration the theory for visual information processing in the early stages of human visual system. It is composed of two main components: the Colour Opponent System (COS) and the Chromatic Segmentation System (CSS). The structure of the CSS architecture is based on BCS/FCS systems, so the proposed architecture maintains the essential qualities of the base model such as illusory contours extraction, perceptual grouping and discounting the illuminant. Experiments performed show the good visual results obtained and the robustness of the model when processing images presenting different levels of noise. © 2008 Springer-Verlag Berlin Heidelberg.