Font Size: a A A

The Research Of Digital Image Watermarking Algorithm Based On Discrete Wavelet Transformation

Posted on:2008-05-05Degree:MasterType:Thesis
Country:ChinaCandidate:F ZhengFull Text:PDF
GTID:2178360215489590Subject:Computer application technology
Abstract/Summary:PDF Full Text Request
The daily life of people are becoming more and more digitalized and networked with the rapid development of computer and internet. Spread of a large number of private data and popularity of electrical commerce made the traditional method of copyright protection and technology of data security face great challenge and difficulty which isn't conquered easily. Digital watermarking is proposed in such situation. Digital watermarking is a new and practical technology of information hiding and copyright protection. It has caused great concern and has been developing rapidly in recent years.The major work in this dissertation is listed as follows:1. Referring to a large number of domestic and foreign literatures of digital image watermarking, the recent research in digital watermarking technology at home and abroad was summed up.2. The concept, characteristic and the algorithm of Mallat was described. Moreover, the advantage of using wavelet transform and the energy distribution of wavelet was analyzed through the experiment.3. In the dissertation, the algorithm of zerotree wavelet coding proposed by Inoue is improved. A new binary image watermarking algorithm based on the human visual characteristics and embedded zerotree wavelet coding is proposed. It classifies the image and image blocks and sets the different visual weight. According to the important wavelet coefficients and texture masking function,the watermarking is embedded. This algorithm is tested in C++ Builder 6.0. From the Experimental conclusion the algorithm based on HVS and EZW has strong robustness. The algorithm can extract an effective watermarking after the image is processed by general signal.4. A new color image watermarking algorithm based on the run length coding and human visual system is proposed. It protects the security of watermarking using pseudo-random sequence. Compressing the run length coding and combining the multi-resolution and JND threshold value of the human visual system, the watermarking is embedded. The algorithm balances the robustness and transparence of watermarking. This algorithm is tested in C++ Builder 6.0.The experiment results show that this algorithm is robust against image processing and attack from JPEG compression, noise,filtering etc.
Keywords/Search Tags:discrete wavelet transform (DWT), human visual system(HVS), digital watermark, embedded zerotree wavelet(EZW), robustness
PDF Full Text Request
Related items