| With the rapid growth of Internet technologies and extensive use of multimedia data, the enforcement of multimedia copyright protection becomes an important issue. Digital watermarking is viewed as an effective way for intellectual property right protection of multimedia data.This paper is focused on images watermarking algorithms. Watermark casting operates in frequency domain, embeds pseudo-random sequence of real number in selected coefficients without image degradation. A blind watermark retrieval technique that can detect the embedded watermark without resorting the original image is proposed, so that the proposed techniques represents a major improvements to those relying on the comparison between the watermarked and original image, which is not practical for a huge image database. All process, include watermark embeding, extracting and verification are simulated in MATLAB environments. Experimental results show that the embedded watermark is robust against various signal processing and compression attacks.The watermark systems in our paper are mainly based on Discrete Cosine Transform (DCT) and Discrete Wavelet Transform (DWT). In DCT-based system, select mid-frequency as the watermark carrier for the consideration of the tradeoff between the imperceptibility and robustness requirements. In DWT-base system, a couple of significant wavelet coefficients in detailed subbands are selected to embed watermarks by first level decomposing. A better image quality (imperceptibility) can be gained by exploiting human visual system.Compared with DCT-based system, DWT-base watermark system is an adaptive system. As the multi-resolution property of DWT enable DWT-based watermark system combine the statistic characteristic of image with human visual system, make it possible to find an optimal system.Besides, Using proposed system test the parameters which work on the system performance, such as embed position, embed strength, watermark structure, image capacity and HVS etc. |
PDF Full Text Request