Contourlet Transform Based Digital Watermarking Resisting2D-3D Conversion

With the continuous development of information technology, multimedia technology has also been developed rapidly.3D video has attracted more and more attention and research, along with "Avatar" and other3D films shown in theaters fiery. The research on the protection of3D technology has also been pushed to the peak of3D multimedia.3D video production sources are divided into three categories:The first category is using two color cameras to capture the videos, to represent the left eye and the right eye videos respectively; the second category is using one color camera to capture the video, and the corresponding depth information is acquired by a dedicated depth camera; the last category is using only one color video to estimate the corresponding depth information through some unique video processing algorithms, which makes the2D-3D conversion urgent need for content and copyright protection, and therefore the digital watermarking algorithm that can be used for2D-3D conversion emerged.2D-3D conversion requires the center image and its corresponding depth image. After the2D-3D transformation, the center image and its corresponding depth image will form two virtual images with certain parallax. The two virtual images can simulate the images from the human eyes, which can form the three-dimensional feeling. And this transformation is usually the Depth Image-Based Rendering technology, which is also called DIBR technology. In the2D-3D conversion process of the digital watermark, the study on DIBR technology has the two-fold meaning. Firstly DIBR image conversion technology can be regarded as a technique as it can provide another form of media on the basis of the existing media directly. Secondly DIBR can be seen as a form of attack, and the resistance for DIBR transform can be used to evaluate the robustness of the watermarking scheme. Therefore, the study of DIBR and digital watermarking technology has two meanings, which are the protection of traditional media and the protection of the generated new media by the conversion.2D-3D conversion brings two main issues of the video, which are spatial synchronization and hole. Firstly the space synchronization can be regarded as the approximate translation transform. The existing studies indicate that the process of DIBR can be considered as horizontal movement approximately and the Contourlet transform is invariant to such translation. That makes it have the ability to resist the approximate horizontal movement of DIBR transform to the watermark information effectively. Secondly the holes lead to the lack of information. The quantization-based watermarking algorithm on zigzag scanned DCT coefficients is helpful to extract the watermark information blindly, and its robustness can effectively deal with the hole issues. Taking into these two issues account, we have carried on a preliminary exploration on the digital watermarking of2D-3D conversion.Simultaneously, we also has carried on several experiments of3D digital watermarking system on different situations and the experimental results show that our proposed algorithm based on Contourlet transform can resist2D-3D conversion and has better robustness and anti-attack capability.