Font Size: a A A

Research Of Digital Watermarking Algorithms Based On Wavelet Transform

Posted on:2007-03-12Degree:MasterType:Thesis
Country:ChinaCandidate:F F HuoFull Text:PDF
GTID:2178360182977798Subject:Signal and Information Processing
Abstract/Summary:PDF Full Text Request
With the rapid growth of information technology, the emergence of e-commerce and the vast of commercial multimedia business have made the media security a very important issue. To develop an effective digital media protection method, digital watermarking attracts much more attention. The robust watermarking techniques for copyright protection and information hiding are investigated in this paper with emphasis on digital watermarking algorithms based on wavelet transform.Firstly, an algorithm for digital video watermarking is proposed based on binary operation in 3D wavelet transform domain. For the sake of convenience to extract the watermark image, a blind detection and restoration algorithm is presented, which does not depend on the original video. It is worth to note that the proposed watermarking embedding and extracting algorithm uses only four image frames for performing wavelet decomposition per operation with lower memory demands and short delay. Hence, it fairly suits for watermark online embedding and extracting and real-time implementation with hardware. The experimental results illustrate that the proposed watermarking algorithm is robust to video degradation and distortion attack, such as Gaussian noise, MPEG coding, frame averaging, frame dropping and frame swapping.Secondly, through the analysis of the correlation of the wavelet coefficients, a novel watermarking scheme is presented based on the triplets of wavelet coefficients in the three lowest detail subband images. The embedding algorithm is based on the correlation of the wavelet coefficients, which leads to very small distortion and guarantees the quality of the watermarked image. Both theoretical analysis and experimental results illustrate its efficiency.Finally, the paper is concluded, and some further research issues are presented.
Keywords/Search Tags:Digital watermarking, 3D wavelet transform, Real-time embedding, Blind detection, Triplet
PDF Full Text Request
Related items