Research Of Fragile Digital Watermark Algorithms Based On Image

The rapid development of computer network technology and digital multi-media technology has brought about a slew of issues with regard to information security. The fact that digital products can be easily altered, pirated and illicitly distributed has confronted us with a new set of challenges. The verification of the integrity and authenticity of digital products is a problem that must be solved and it must be solved now. As an important branch of information hiding, digital watermarking technology has played an irreplaceable role in resolving IPR disputes and verifying the integrity and authenticity of digital products. Fragile watermarking, as part of the digital watermarking technology, is an academic hotspot for researchers and scholars around the world. An in-depth analysis of fragile watermarking technology is presented in this document.After briefly exploring the existing theories and algorithms about digital watermarking, image scrambling techniques are introduced in this thesis. The advantages and disadvantages of Asia affine transform are analyzed and a fragile watermarking algorithm based on LSB and image scrambling is presented in this paper. Since both the original image and watermark image in the algorithm were scrambled, a high level of invisibility of the watermark image is achieved. Furthermore, the number of Asia affine transforms applied to both watermark image and original image is regarded as the private key, which has improved the security of the whole algorithm.The second fragile watermarking algorithm proposed in this paper based on DCT and RSA, in which the features extracted from the original image is used as the watermark signal. First the original image is divided into blocks of predetermined size and DCT is applied to each and every block. DCT coefficients are viewed as the image feature and watermark signal can be obtained by applying Hash function and RSA cryptographic technique to the extracted feature, namely the DCT coefficients. Since the watermark signal is obtained from the original image, the proposed algorithm is self-adaptive and no extra memory is required for watermark information. The source of watermark signal and the encrypting processing have significantly enhanced the security of the algorithm and the dividing of the original image is very effective for tampering localization.After exploring the foundmentals of wavelet transform, lifting technique of wavelet transform and chaotic sequence, a novel fragile watermarking technique based on integer wavelet transform and chaotic sequence was presented in this document. The watermarking embedding process was determined by chaotic sequence and coefficients of integer wavelet transform. By so doing, the complexity and security of the proposed scheme have been substantially improved. Theoretical analysis and experimental results both show that the proposed scheme is highly fragile to changes to the watermarked image. Moreover, the medical and forensic applications of the proposed algorithm are also explored in this thesis.