A Morphological Approach for Granulometry with Application to Image Denoising

摘要

Granulometry formalizes the intuitive geometric notion of a sieving process. It was initially set oriented to extract size distribution from binary images, and has been extended to function operators to analyze and extract texture features in grey-scale images. In this paper, we study and establish granulometry with respect to grey-scale morphological operators based on fuzzy logic. We discuss applications of the granulometry in image analysis. A numerical experiment shows that granulometry is a powerful tool for image denoising for image analysis and processing.