Digital Watermarking Algorithm Based On Human Vision Wavelet Transform

With the development of the digital technology and internet, some digital products such as image,audio,video and so on have been published as the form of network, the product of their copyrights has become the problem which need to be solved immediately. The security protection of the multimedia data itself has special requirement, such as they require the encrypt algorithms have low calculate intensity, the security of encrypting can not depend on the complication of the algorithms, the security protection to data can not affect its availability and so on. The protection of the conventional cryptology is not very effective, as the important instrument for the protection of copyright, the newly developed digital image watermarking technology has become a hotspot in the area of image protection technology recently, while the theory of wavelet analysis is the front filed which is highly concerned by the science and technology at present. It has proved by reduplicate practice, among the multitudinous digital watermarking algorithms wavelet transform algorithms has more speciality than the other algorithms, it can well realize the protection and authentication to the digital products.In this thesis, fist expatiate the need background of the digital watermark and the present study status domestic and abroad, introduce the basic principles systematically, analyze the characteristic, classes and application area of the digital watermarking, at the same time introduced the basic theory about DWT, in the basic theory introduced the develop history of wavelet analysis, the basic definition multi-resolution analysis of wavelet, Mallat algorithm, wavelet transform characteristic and its application in the image processing. Then some image-scrambling algorithms were introduced, in the end introduced DWT digital watermark algorithm combines with the human visual system and the improved algorithm.The main idea of the use of wavelet transform in image processing is that signal is expressed on time field and scale domain, decomposition dimensions of domain being to should in different frequency range. The low frequency tape has kept the main information of the image and it decide the human apperceive degree, but the edge and the texture features of the image are general to concentrate in the high frequency tapes, if we add the watermark into the low frequency tape, it can lead to the reduce of the visual effect of image, if we add the watermark into the high frequency tapes then the information of the high frequency tape will be lost after general image processing, and the robustness of the digital watermarking will be reduced. Considering the above all, we should add the watermark into the intermediate frequency tapes, then the invisibility will be assured and the robustness of the algorithm will be enhanced.In order to add more information of the watermark in the case that the quality of the original image will be kept, the watermarking and the embedding algorithms in low and middle domains, associated with the human visual system and image sharing technology, are proposed. The result of the experiment indicates that the algorithm improved can balance the invisibility the steadiness and its capability, the algorithm can assure the robustness to familiar image process such as noise contamination, compression and cropping, it is a feasible algorithm and it make an active sense to the present wavelet digital watermarking algorithm.