Research and application of a new computational model of human vision system based on ridgelet transform

The properties and limitations of the human visual system (HVS) can be utilizing in many image processing, such as image compression, watermarking and image quality assessment. This paper addresses the issue of how to establish the computational model of HVS in image ridgelet transform domain. Utilizing the multi-channel spatial-frequency decomposition, we proposed a new visual model in ridgelet domain by taking into account the local band limited contrast, the frequency sensitivity and masking effecting of HVS. The proposed visual model can exactly evaluate the just noticeable distortion (JND) tolerance of HVS. Finally we demonstrated the performance of our model with the application in the field of image quality evaluation and digital watermarking. A perceptual criterion of image quality evaluation is proposed and a robust watermark in ridgelet domain is presented. Lots of experiments are shown the new computational model of HVS is very useful in such applications.