A working spatio-temporal model of the human visual system for image restoration and quality assessment applications

This paper describes a spatio-temporal model of the human visual system (HVS) for video imaging applications, predicting the response of the neurons of the primary visual cortex. The model simulates the behavior of the HVS with a three-dimensional filter bank which decomposes the data into perceptual channels, each one being tuned to a specific spatial frequency, orientation and temporal frequency. It further accounts for contrast sensitivity, inter-stimuli masking and spatio-temporal interaction. The free parameters of the model have been estimated by psychophysics. The model can then be used as the basis for many applications. As an example, a quality metric for coded video sequences is presented.