Research On Robust Image Watermarking Algorithms Based On Curvelet Transform

With the development of the multimedia technology and Internet, digital watermarking becomes the research hotspot because it is an effective way to protect the safety of digital multimedia information. Digital image watermarking is to embed a logo (image, text) of copyright information into the original image to achieve the effect of copyright protection. Due to transform domain digital image watermarking has good invisibility and robustness, it becomes popular.Curvelet transform is sparse than Wavelet transform, it can gives more efficiently represent to edges and other curves in images. The first generation of curvelets was designed base on ridgelet transform theory in 1999. It needs subband decomposition, smooth, renormalize and ridgelet transform in its realization, the process is complex and has high redundancy. In 2002, the second generation of curvelets which is also named Fast Discrete Curvelet transform (FDCT) was born. It is anisotropic with strong direction like the first generation of curvelets. But its implementations are simpler, faster and far less redundant than the earlier implementations. In recent years, curvelets have been widely used in image processing field such as image de-noising, image fusion, image enhancement, image watermarking and so on. In this paper, the robust digital image watermarking algorithms are emphatically studied. The main works are as follows:(1) Analyses the study background, the significance and the research status of digital image watermarking, introduce the emergence and development of curvelet, and make further introduction to the conception, system model, basic feature, classification and application of digital image watermarking. The typical technologies and evaluation criteria of image watermarking system are also introduced.(2) The basic knowledge of multiscale geometric analysis is briefly introduced; Then the first generation and the second generation of curvelets and its implementations are summarized. On this basis, analysis the characteristics of the curvelet through examples, and summarize the key techniques and advantages when curvelet transform is used for digital image watermarking.(3) A new multiresolution watermarking algorithm based on fast curvelet transform is proposed. It embed multiresolution watermark of a binary image into the same resolution curvelet coefficients of the carrier Image. Experiment results show that the scheme is simple and effective, and it is invisibility and robust against various signal processing such as JPEG compression, noise adding, filter, cropping, and so on.(4) A digital image watermarking algorithm based on fast curvelet transform is proposed. Firstly, the carrier image is decomposed by fast curvelet transform, and, the watermarking image is scrambled by Arnold transform. Secondly, the binary watermarking image is embedded into the medium frequency coefficients according to the human visual characters and curvelet coefficients. Experiment results show that the proposed algorithm has good performance in both invisibility and security. It also has good robustness to the noise, cropping, filtering, JPEG compression and other attacks.