Performance comparison of different denoising methods for sonar images

Abstract

This paper focuses on noise reduction of sonar images, which is one of the most important tasks in the pre-processing step of sonar image processing, and compares the performance of different noise reduction techniques. Noise reduction is an important pre-process for the segmentation and classification of sonar images, because the success of subsequent processes is directly depending on the success of denoising. In this paper, the performance of different noise reduction methods for sonar images is evaluated. For this purpose, phase preserving denoising which has not been applied to sonar images so far has been evaluated and shown to provide good results, in addition to commonly known Wiener filtering and Discrete Wavelet Transform based denoising approaches.