Watermarking Algorithm Based On FCM And Wavelet Transform Analysis

With the booming development of information industry, the digitalization of information media and exceptional growing of computer networks, we can get digital information conveniently and quickly. However, the digital multimedia can be easily exploited by pirates. In the open computer network environment, it has become much urgent for us to provide efficient copyright and security protection means for the digital information products. So the digital watermarking technology has been put forward for this in recent years. Digital watermarking is a key way of copyright protection, one of which image as carrier has attracted many researchers.Wavelet analysis is the newest implement for space-frequency analysis at present, and it has been widely used in image processing. As the characteristic of human visual system(HVS) has the compatibility with the multi-resolution analysis characteristic of wavelet transform. The paper taking static image as fundamental research object, thorough research has been conducted to the digital watermarking algorithms in wavelet domain, combined with the human visual system(HVS) characteristic and Fuzzy C-Mean(FCM) clustering, two watermarking algorithms in wavelet domain are proposed as below:1,This paper proposes a watermarking algorithm based on wavelet transform and Fuzzy C-Mean(FCM) clustering. Watermarking embedding locations are identified on the wavelet domain coefficient by applying FCM clustering algorithm based on the human visual system. And the binary image is embedded according to the image local correlation. In order to improve the security and robustness, the watermark is scrambled before embedded. Experiment results show that the watermarking scheme is invisible and the algorithm is strongly robust for the common image processing techniques such as JPEG compression, cropping, salt and pepper noise addition, low-pass and median filtering and so on.2,The paper presents a new multiple watermarking algorithms based on texture features of wavelet domain. Both the gray watermarking and the binary watermarking and binary watermarking are embedded in the host image, which can better protect the original copyright. A host image is decomposed with two level wavelet transform, then the middle-frequency coefficients of wavelet domain based on texture features are classified and coefficients of gray watermarking are embedded as binary intensity factors. This algorithm can get the watermarked images with perceptual invisibility and strong robustness. In order to improve the security of the watermarking algorithm, the two watermarking are scrambled before embedded. Experiment results show that the algorithm is strongly robust for the common image processing techniques such as JPEG compression, cropping, salt and pepper noise addition, low-pass and median filtering and so on...