Reversible Watermark For Color Images Based On Difference Expansion

With the growing popularity of the Internet, the problem that illegally accessed and copied digital products becomes more serious. Following this, the copyright of digital products and the integrity protection has been the focus gradually. The mature program to protect multimedia information security is cryptography, however, the cryptography protects copyright by the way of digital signatures which need additional Authentication information that Limited the scope of application of cryptography. Digital watermark is the way which embedded a number of authentication information in the digital carrier directly, and will not affect the use value of the carrier, the embedded information can be used to identify the copyright owner of the legality of digital carrier. Digital watermarking is widely used in digital media copyright protection, tamper detection, tamper localization and content of certification.There are some digital information, such as medical digital images and multimedia information for legal evidence, a slight modification may lead to serious consequences. Such images need extract the watermark and completely recover the original image - the reversible digital watermarking technology. Color images has become the mainstream of contemporary, based on the reversible watermark grayscale has been unable to meet demand, Therefore, this paper focuses on the study of reversible color image watermarking.This paper describes the background of digital watermarking, significance and research status. Then analyzing the basic theory of digital watermarking and describes the characteristics of reversible watermarking and attacks against the reversible watermarking methods and performance evaluation.Secondly, described the method of digital image storage and two main categories of image reversible watermarking algorithm, then introduced two reversible watermarking algorithms which based gray scale image, they are difference expansion and prediction error expansion.Subsequently, analyzed the relevance of color images, and combine the two grayscale watermarking algorithm, proposed a color image reversible watermarking algorithm. The algorithm used the correlation of color images to get small and centralized difference, and combined with histogram shifting embedded data in peak point of the Positive and negative Axle of the histogram, then used 0 point to reduce the pixel which need to shift. Avoided overflow by preset pixel value, and identified adjusted pixels by adjustment information, Therefore, the location mop has be replaced. Experimental results show that the algorithm can take full advantage of the correlation of the three color components of color image, Guaranteed the image quality and providing a high embedding capacity, compared with other Reversible watermarking algorithm, overall performance of proposed algorithm has obvious advantages.Finally, the lacks of the proposed algorithm were analyzed and discussed the future development trend of reversible watermark.