Region Based Segmentation Using Dual Tree Complex Wavelet Transform and Rotated Complex Wavelet Filter

Abstract

In this paper, we have proposed a novel approach for segmentation of textured images using combination of dual tree complex wavelet transform (DT-CWT) and dual tree rotated complex wavelet filters (DT-RCWFs). DT-CWT is used because of its properties such as shift invariance, good directional selectivity, limited redundancy and efficient computation. RCWF sets provide important complementary information to the DT-CWT filter set by extracting texture features in 6 different directions which are 45° apart from decomposition directions of DT-CWT. Mean shift algorithm is used along with fuzzy c-means (FCM) to make segmentation automatic. The proposed technique is tested on a range of textured images including composite texture images, synthetic texture images as well as real scene images.