Implementation of Proposed Threshold for Despeckling in Stationary Wavelet Domain

Abstract

Medical images are prone to different types of noise. Such types of noise corrupted images leads to incorrect diagnosis. Hence, removal of noise is a prerequisite in medical imaging modality. Speckle noise is widely found in coherent medical images, like in Ultra Sound images and Optical Coherence Tomography images.In the preprocessing stage, the noise present in the medical image has to be removed while preserving the edge information and other structural details of the image. Relevant denoising technique has to be chosen based on the nature of the medical image. This research is focused on design of algorithms for speckle denoising of Ultra Sound images and Optical Coherence Tomography images in stationary wavelet domain. Standard speckle filters in wavelet domain were analyzed and compared with the proposed method. Results obtained proved that the proposed method was able to remove speckle noise while preserving better edges.