New Method for Performance Evaluation of Grayscale Image Denoising Filters

摘要

Combining noise removal and detail preservation is a challenging issue in the design of image denoising filters. In this letter a new full-reference method for objective evaluation of such features in grayscale pictures is presented. The proposed approach is based on the classification of the filtering errors in two classes that respectively take into account the amount of noise still present after processing and the actual distortion produced by the filter. Results of computer simulations show the advantages of the proposed method with respect to other measures of filtering performance that are available in the literature.