Mca Neural Network Applications In Digital Watermarking

Nowadays, with the fast popularity of the computer technology, many and many artists and writers prefer to record and restore their works directly by digital way. Moreover, the rapid development of internet also provides great convenience to transmit such digital works. All these merits make the digital works, compares to traditional works, have great advantage in create and broadcast such works. However, it is very easy to copy the digital works, which is also the main reason that digital works could broadcast fast. But such property is easily misused by infringers. Right now, the pirate has already been the greatest threat to digital industry, which, to some degree, has greatly baffled the development of the digital industry. For the owners of the copyright of the digital works, anti-copy is in need desperately. In one word, the protection of the copyright of the digital works is not only a matter of legal problem, but also a technology problem.Digital watermark is a newly proposed method to protect the copyright. It takes advantage of the redundant data and the randomicity to insert the copyright information into the digital work itself. It is a new technology to protect the copyright of the digital works. Digital watermark could label and identify the information of the creators, owners, and the consumers of the digital image, video and audio record. It could also trace the illegal release of the digital works. To sum up, digital watermark is an efficient technology method to protect the copyright of the digital works.Meanwhile, in many applications of the signal process, the extraction of the minor component of the input signal has great meaning and merits. As an important tool of signal process and data analysis, MCA (minor component analysis) has been widely used in beamforming, frequency estimation, curve/surface fitting, computer graphics and many other engineering fields.In this paper, we for the first time apply the MCA neural networks to the field of digital watermark, and propose two new digital watermark algorithms: digital watermark based on RSA and MCA neural networks, and the digital watermark based on private key and MCA neural networks. We introduce in details of the two kinds of algorithms, and make a test of the performance of the visual quality and anti-attacks of the two algorithms. Finally, we summarize the properties of the two proposed digital watermark algorithms.We also introduce in details the design and implementation of the digital seal system. The purpose of the digital seal system is to provide a contract assignment service through Electronic Commerce. It consists of digital seal management sever, digital seal management terminus, digital seal assignment terminus, and digital seal client, etc.. Based on this system, my main work is to implement and integrate the new digital watermark algorithm based on RSA and MCA Neural Networks into the digital seal system.