A Test Platform for the Security and Quality of Video Watermarking Content Protection System

In this paper we present a architecture of a video watermarking evaluation system. Based on the disadvantages of the existing watermarking benchmark software and basic principles, a comprehensive, software and hardware combined test platform with objective and subjective test methods is proposed to solve the real video watermarking application, whose feature is it can solve the deficiency by the pure software testing. Considering the weakness of the pixel-based distortion metrics and the existing JND-based perceptual quality metric, an improved PSNR, namely APSNR is proposed to evaluate the performance of the quality of the video watermarking schemes, which has three advantages: automated, objective measure, combined human visual system (HVS). Simulation results show the good performance of the proposed schemes.