Distance transform measure based on edge-region information: An algorithm for image quality assessment

摘要

In this paper we consider the problem of objective assessment of image quality. The Mean Structural Similarity Index (MSSIM) have been proposed to assess image quality, inspired by comparing the structures of the distorted and the reference images and presents image distortion as a combination of three factors: loss of correlation, luminance distortion, and contrast distortion. We identify some limitations in this assessment method, and we propose an enhanced measure that corresponds more closely to visual judgment.