A comparison of image quality models and metrics based on human visual sensitivity

This paper considers the use of image quality metric for still image compression systems comparison. Considering the fact that the conventional PSNR cannot sufficiently reflect the result of subjective assessment, other quality measures have been considered to design the variable bit-rate coders. Indeed, distortion measures are developed for the purpose of predicting the subjective quality of pictures. Subjectively relevant distortion measures that mirror viewers' assessments of picture quality would make the task of designing and optimizing coding schemes considerably easier. One significant problem associated with these measures is that there is little information on how these metrics perform in comparison to each other. The purpose of this paper is to provide a rigorous evaluation of three metrics to assess the quality of compressed images. The compression system used in this evaluation is the classical JPEG coder. Both objective and subjective tests were performed on a 50 natural image data base with a panel of experimented and non experimented observers. The results show that these metrics are highly correlated with the subjective quality grading but also depends on the complexity of the images under study. They can be improved by integrating all masking effects and typical artifacts of the compression method. Their combination as a multidimensional quality measure in which each dimension related to a properly identified artifact, should be based on a subjective evaluation.