Intelligent Image Watermarking based on Handwritten Signature

Abstract

With the growth of digital technology over the past decades, the issue of copyright protection has become especially important. Digital watermarking is a suitable way of addressing this issue. The main problem in the area of watermarking, is the balance between image transparency and resistance to attacks after watermarking, where an increase in either one of them will always cause a decrease in the other. Providing statistical and intelligent methods, is the most common way of optimizing resistance and transparency. In this paper, the intelligent method of genetic algorithm (GA) in watermarking will be examined and also the results of using this method will be compared with the results of a statistical SVD-based method. Also, by combining the issues of watermarking and authentication, a relatively higher security in these two issues can be achieved. In this scheme, the security of watermarking increases through the provision of a new method which is based on the combination of image watermarking with a person's handwritten signature. It must be mentioned that the section of signature recognition is implemented using neural networks. The results from implementing these two methods show that in this area, intelligent methods have a better performance compared to statistical methods. This method can also be used for tasks like passport or national identity card authentication.