Influence of Sorting Measures on Similar Segment Grouping based Denoising Algorithms

Abstract

Denoising is a fundamental challenge in the field of digital image processing and computer vision. Many of the denoising algorithms operate on segments of images in which non-locally distributed similar segments are identified and grouped to form sub-images. The denoising algorithm is applied on these sub-images to recover the actual image. The basis of such denoising algorithm is the existence of non-local self similarity or redundancy of segments within the natural images. But choosing the similarity measures for segment segregation is challenging. An analysis of various sorting measures and its effect on the performance of the denoising algorithms is presented.