Design a system for an approved video copyright over cloud based on biometric iris and random walk generator using watermark technique

Abstract

Copyright is a tool for preventing anyone forged to copy an electronic work from another person and claim that electronic work is referred to him. Since the identity of the person is always determined by his name and biometrics, there is a concern to handle this information, to preserve the copyright. In this paper, a new idea for copyright technology is used to prove video copyright, by using blind watermarking technique, the ownership information is hidden inside video frames using linear congruential generator (LCG) for adapted the locations of vector features extracted from the name and biometric image of the owner instead of hidden the watermark in the Pseudo Noise sequences or any other feature extraction technique. When providing the watermarked vector, a statistical operation is used to increase randomization state for the amplifier factors of LCG function. LCG provides random positions where the owner's information is stored inside the video. The proposed method is not difficult to execute and can present an adaptable imperceptibility and robustness performance. The output results show the robustness of this approach based on the average PSNR of frames for the embedded in 50 frames is around 47.5 dB while the watermark remains undetectable. MSSIM values with range (0.83 to 0.99).