Multichannel Texture Analysis Using Localized Spatial Filters
IEEE Transactions on Pattern Analysis and Machine Intelligence
Unsupervised texture segmentation using Gabor filters
Pattern Recognition
Texture Segmentation using 2-D Gabor Elementary Functions
IEEE Transactions on Pattern Analysis and Machine Intelligence
Hi-index | 0.00 |
Many proposed texture-segmentation schemes are based on a filter-bank model. The filters, henceforth referred to as Gabor filters, have typically been designed empirically. Dunn et al. have recently derived analytical criteria for designing appropriate Gabor filters; they did not discuss how to design filters for general natural textures. This paper presents an algorithm for designing optimal Gabor filters. The algorithm assumes that an image contains two different textures and that prototype samples of the desired textures are given. It uses a decision-theoretic framework, based on modeling a Gabor-filter output as a Rician distribution, for designing optimal filters. To gain more robust results, we also propose a multiple-filter segmentation scheme. Experimental results verify the efficacy of our methods.