Research On Digital Image Based On Wavelet Transform

With the rapid development of multimedia and network technology, all kinds of digital multimedia products, including image, audio and video etc. have been widely distributed and exchanged through network. At the same time, they also offer convenience to some illegal persons who can distort, attack and pirate the achievement maliciously. How to avoid the information of the net being copied and distorted, the watermarking technology emerges. So-called digital watermark embeds certain amount of important information in digital host media. According to different requests, the added information mostly maybe invisible. The digital watermark technology can protect not only the copyright but also the authentication of the content. Now this technology has been a new issue of multimedia information security area.In this thesis, we propose a digital watermarking algorithm based on DWT, which is from the research of multimedia static image. The key work of this paper includes the following several aspects:1. Summary of the digital watermarking is presented from origin, present situation, characters and classifications.2. By comparing several algorithms of digital watermarking, the advantages and algorithms of wavelet transform put forward. Now the digital watermarking algorithm is widely used in JPEG-2000.3. Compared with the traditional algorithm, the new scrambling algorithm is an improvement in the Arnold algorithm. This improvement make the scrambling image good attack-resistant.4. In this paper, we also propose a new watermarking algorithm based on wavelet transform. At first the host image is divided into blocks , which can lower time complexity effectively. Then through wavelet transform to the host image, we embed the watermark. This algorithm makes a scrutiny into the energy distribution of wavelet image and gets the appropriate threshold and coefficient.The experimental results show the effectiveness of the watermarking algorithm based on wavelet transform, which has good invisibility and robustness to malicious attack.