Video Watermarking Algorithm PSO And Wavelet Transform

In the new era of unceasing development of digitization and informatization, a large number of digital video products have sprung up. People are enjoying which bring huge convenience, at the same time, also facing the great challenges of copyright violations. The problems of copyright disputes and security authentication become more and more serious. Video watermarking technology has become an effective means which the world related industries solve the problem of copyright protection. Due to the inherent characteristics and complexities such as the large volume of video data, campaign regional imbalances and video compression standard diversification, its performance requirement is far higher than the static image watermark. The requirements are mainly such as lower complexity, higher robustness, large capacity, and blind extraction, etc. Most of the current algorithms are difficult to achieve the optimum and balance in these respects. Therefore, in this paper there are proposed two kinds of watermarking algorithm based on original video. One improve the robustness and invisibility of watermark through searching the best embedding strength of the watermark by particle swarm optimization and achieving multilevel embedding watermark by the multi-resolution of wavelet transform. The other achieve the blind extraction of watermarking through the stability of the singular value decomposition.To improve the robustness of watermarking algorithm and achieve the blind extraction of watermarking, an adaptive video watermarking algorithm based on3D wavelet transform and singular value decomposition is proposed. In the algorithm, the video sequences are divided into isometric group at length of4frames firstly, and one video series is selected as key frames by key, then,3D wavelet transform is carried out in the luminance components, the sub-blocks are divided from the frequency region, and each is decomposed with singular value decomposition. Finally, the watermark information is embedded into the maximum singular value of the blocks by quantization method. The algorithm realizes can achieve blind extraction, the experiments show that the algorithm has good robustness against the noise attack, frame replacement, frame shear and so on.In order to improve the watermark robustness and invisibility, a video watermarking algorithm is proposed based on particle swarm optimization and wavelet transform. Considering the watermark embedding process and the robustness and invisibility mutually restricted, how to adjust the contradictions between the two is the key point. The video sequences are divided by histogram segmentation at first. Then select luminance components of four continuous video in each scene as the watermark embedding area. Because the watermark information is embedded into the low-frequency part of wavelet transform, it can reduce the invisibility, while it is embedded into the high-frequency part of wavelet transform, it reduces its robustness. Watermark information is embedded in the middle frequency coefficients of3D discrete wavelet transform by multiplication embedding algorithm, Overcome the shortcomings of the watermark embedded in the low-frequency coefficients and high frequency coefficients, embedding intensity of watermarking is searched by particle swarm optimization, too large embedding strength impacts the invisibility of the watermark, conversely reduces its robustness. The simulation results show that the method not only has invisibility, but also has good robustness against noise attack, frame average and MPEG-2compression attack test etc, the results indicate that the algorithm has good robustness, it is to achieve the objective of the comprehensive optimization of the invisibility and robustness.The algorithm that3D wavelet transform is combined with singular value decomposition, not only improved the robustness of the watermark, but achieved blind extraction. The algorithm that the best watermark strength is searched by particle swarm optimization algorithm, and watermark information is embedded in the middle frequency coefficients of3D discrete wavelet transform by multiplication embedding algorithm, not only improved the robustness, but was good invisibility.