Quality assessment for a perceptual video encryption system

Abstract

Perceptual video encryption, in which a user can still obtain some visible video contents (but at a very annoying quality) even without knowing the encryption key, is becoming more and more interesting in video applications such as video-on-demand (VoD) and pay-TV. Since not all visual information is corrupted, the quality assessment methods that are used in a total video encryption system (completely confidential) may not be suitable in this case. In this paper, we study the performance evaluation for a perceptual video encryption system we developed recently by using alternative transforms [1]. To this end, we consider both visual quality based assessments and quantitative assessments. Two quantitative metrics, namely PSNR and SSIM, are used to evaluate our perceptual encryption system quantitatively, while the SSIM-based one is found to be very consistent with the visual quality assessment.