Research On Robust Video Watermarking Algorithm Based On Human Visual System

Video watermark is an important mean of video copyright protection, it achieve the purpose of authentication through inserting the specific information of logo copyright owner into video sequence. According to the video watermarking embedding areas, video watermark is divided into airspace video watermarking, transform domain video watermarking and compressed-domain video watermarking. Airspace video watermarking has high robust, so that can be effective against a variety of targeted attacks; compressed-domain video watermark makes the video can be shared and stored in the transmission channel more efficiently and easily, it becoming the most notable video watermarking algorithm.The key technologies of video watermarking include watermarking embedding algorithm, watermarking detection and extraction algorithm. These algorithms affect some performance measurement indicators of video watermarking system, such as covert, security, robustness and so on. Revolved round the key technologies of video watermarking, this paper carries out research as follows:(1) For the robustness deficiency of the existing video watermarking algorithms, combined with the human visual system,a new video watermarking algorithm of moving object detection based on Daubechies complex wavelet transform is put forward. Firstly, the original image is divided into 8 x 8 sub-blocks and each sub-block is decomposed by Daubechies complex wavelet. In order to solve the problems of ghosts, shadows and noises, frame difference method based on approximate median filter is used to detect the moving targets and mark the movement area more accurately. Then the watermarking is preprocessed by using Arnold scrambling and chaotic encryption to enhance the security of the watermark. Finally the moving area and the area with highly texture complexity are chosen as the embedding area of video watermarking. The results show that the algorithm has the property of imperceptibility and good robustness against the common video attack. The performance comparison with three existing algorithms also reflects the superiority of the proposed algorithm.(2) Towards the shortcoming of video watermarking ability in the resistance to the recompression and high time complexity, this paper presents a novel video watermark embedding scheme based on chrominance components for H.264. For the deep analysis of the recompression for H.264, embedding the video watermarking in the chrominance components has effectively solved the asynchronous issue owing to the mode conversion during the recompression. Moreover, the video watermarking is embeded by modulating the non-zero quantized DCT coefficients of chrominance components based on coefficients’ parity. The algorithm reduces the time complexity to satisfy the real-time application such as broadcast monitoring, webcast and so on. Compared with the existing algorithms in the resistance to the recompression, the algorithm greatly improves the quality of the reconstructed video images, has much less value of BIR and BER for the extracted watermarking image and good robustness against the recompression.