The Researches Of Digital Image Watermarking Based On Chaotic Systems

In recent years, substantial advantages in multimeadia and web technology have facilitated the production and distribution of image content. However, the ease of digital image editing and copying also enables unauthorized tampering and use of the image content. The design of novel techniques for image authentication and copyright protection has become an important necessity. The chaotic system based digital image watermarking has been proposed as a good means to complement cryptography by using chaotic systems to embed a message within the image content. The message can be used to provide a certain degree of protection to the image from malicious attack or illegally made copies. This dissertation concentrates on the chaotic system based digital image watermarking, and the main contribution can be summarized as follows:(1)Considering the digital image authentication, a fragile chaotic watermarking scheme is proposed. The watermark is generated by quantized wavelet coefficients of the cover image, and the randomicity of the chaotic sequence is used to encrypt the watermark and select the embedded layer. The proposed scheme can accurately detect and locate the place being tampered in the watermarked image. Moreover, it could distinguish the tamper of image between high bits and low bits.(2)Considering the digital image copyright protection, a robust chaotic watermarking scheme is proposed to improve the robustness of common attacks, especially JPEG compression. The chaotic watermark is generated from the iteration of the skew tent map, and the embedding procedure is conducted by adjusting the DCT low-frequency coefficients in the concept of quantization module. The embedded watermark can be almost fully extracted from the JPEG compressed images with very high compression ratios.(3)In order to guarantee the authentication and copyright protection of digital image, a scheme based on chaotic mapping and singular value decomposition (SVD) is proposed. Two chaotic maps are employed to scramble the significative watermark matrix, and the watermark is embedded into the first column of U component in the image SVD domain. There is good resistance of the proposed scheme against general image processing. Furthermore, it can detect and locate the place being tampered in the watermarked image, and the extracted watermark can still be easily identified.