A Perception Based Image Quality Assessment Model

Three important pieces of information are drawn out from the process of subjective image quality estimation.For each of the information,a model is established and also a function is found out to simulate its work on the human visual system.Then a mult-i linear regression analysis is used to integrate these functions into an equation and used as our perception based image quality assessment model.Experimental results show that our estimation closely approximates human subjective tests such as mean opinion score,with high Pearson and Spearman correlation coefficients of 0.966 and 0.949.It is a significant improvement over some typical objective image quality estimations such as peak signal-to-noise ratio.