HVS-based Image Quality Assessment for Digital Cinema

This paper proposes an approach to improve the performance of peak signal-to-noise ratio (PSNR) and structural similarity (SSIM) for image quality assessment in digital cinema applications. Based on the particularities of quality assessment in a digital cinema setup, some attributes of the human visual system (HVS) are taken into consideration, including the fovea acuity angle and contrast sensitivity, combined with viewing conditions in the cinema to select appropriate image blocks for calculating the perceived quality by PSNR and SSIM. Furthermore, as the HVS is not able to perceive all the distortions because of selective sensitivities to different contrasts, and masking always exists, we adopt a modified PSNR by considering the contrast sensitivity function and masking effects. The experimental results demonstrate that the proposed approach can evidently improve the performance of image quality metrics in digital cinema applications.