Texture classification based on Markov modeling in wavelet feature space

Abstract

One difficulty of texture analysis and classification in the past was the lack of adequate tools to characterize textures over different scales. Recent developments in multiresolution analysis, such as the wavelet transform, promise ways to overcome this difficulty. In this paper, we present a texture classification methodology that is based on a stochastic modeling of textures in the wavelet domain. The model captures significant intrascale and interscale statistical dependencies between wavelet coefficients, which are typically disregarded by wavelet-based statistical signal processing techniques. It provides an accurate multiscale texture representation and underlies a highly discriminative texture classification algorithm.