Unsupervised evolutionary segmentation algorithm based on texture analysis

Abstract

This work describes an evolutionary approach to texture segmentation, a long-standing and important problem in computer vision. The difficulty of the problem can be related to the fact that real world textures are complex to model and analyze. In this way, segmenting texture images is hard to achieve due to irregular regions found in textures. We present our EvoSeg algorithm, which uses knowledge derived from texture analysis to identify how many homogeneous regions exist in the scene without a priori information. EvoSeg uses texture features derived from the Gray Level Cooccurrence Matrix and optimizes a fitness measure, based on the minimum variance criteria, using a hierarchical GA. We present qualitative results by applying EvoSeg on synthetic and real world images and compare it with the state-of-the-art JSEG algorithm. © Springer-Verlag Berlin Heidelberg 2007.