Abstract

This chapter proposes a wavelet data hiding scheme for video authentication and ownership protection. A watermark in the shape of a logo image will be hidden. In this research, a discrete wavelet transform (DWT) process is implemented using orthonormal filter banks, where the Y components of the YUV color space of the video frames are decomposed using DWT, and a watermark is inserted in one or more of the resulting sub-bands in a way that is fully controlled by the owner. Then, the watermarked video is reconstructed. The filters used for the DWT decompositions are randomly generated to increase the security of the algorithm. An enhanced detection technique is developed to increase the reliability of the system. The overall robustness of this scheme is measured when common attacks are applied to the test videos. Moreover, the proposed algorithm is used with the high-efficiency video coding (HEVC) technique to examine the whole performance. Furthermore, a selective denoising filter is built to eliminate the effect of the noise. The simulation results show that the proposed algorithm achieves well under both the visual and the metric tests. Moreover, it performed well against intentional and unintentional attacks. The average normalized correlation achieved is 97%, while the mean peak signal-to-noise ratio (PSNR) is 45 dB.

Keywords: watermarking, filter banks, data hiding, pyramid transform, video coding, video attacks
