A comparative analysis of diffusion-based and wavelet-based approaches for despeckling of ultrasound medical images

Abstract

In this paper, Speckle Reduction Anisotropic Diffusion (SRAD) and the modified SRAD approach from spatial domain theme and Wavelet-Transform-based Thresholding Neural Network (WT-TNN) approach from transform domain theme have been extensively analysed based on the speckle suppression and edge preservation of ultrasound medical images and standard images. The results of such comparative analysis of two themes indicate that predicting the number of iteration required for optimum performance is a major concern for SRAD and MSRAD, whereas in case of WT-TNN approach, more despeckling time is a major limitation although it gives better speckle suppression than SRAD and MSRAD. Copyright © 2011 Inderscience Enterprises Ltd.