It is often considered that the implementation of communication security can be done by encryption (CRYPTOGRAPHY). First, the multimedia data file is first encrypted into secret text, so that the illegal attackers that appear during the network pass cannot obtain confidential information from ciphertext, which achieves the purpose of copyright protection and information security, but this does not completely solve the problem. On the one hand, the encrypted document is due to its incomprehensibility hinders the propagation of multimedia information; the multimedia information is easily causing curiosity and attention to the attacker after encryption, and the possibility of being cracked, and once the encrypted file is broken The content is completely transparent. Password has always been considered to be the main information security means in communication research applications and is greatly valued until this situation has changed in recent years. People try to hide the secret information in the ordinary documentation to track the evidence of infringement and provide evidence of legal protection, which is called information hidden or more strictly called information camouflage (Steganography). The word Steganography comes from ancient Greece, meaning that useful or important information is hidden in other information to cover up. Unlike encryption, encryption system protects the content of information, and the information camouflage masks them.
Another aspect of an extremely close relationship with the information camouflage is a research on digital watermarking technology. The information camouflage is hidden in the host data and sends it out. It has a basic assumption that third parties do not know the presence of hidden data, and primarily applied to the secret communication of point-to-point points. Therefore, it does not have robustness requirements, or once the data passed is changed, the embedded information will not be recovered. Digital watermark has robust requirements, that is, it can resist third-party attacks or normal, standard data operations and transformations. In other words, even if an attacker knows that the data passed includes hidden important information, it is not possible to extract it, or the watermark cannot be destroyed without serious damage to the host data. Therefore, digital watermarks and information camouflage is two aspects of a technology complementary, and it is also a frontier hot direction of international academic research, and can provide a potential effective solution for issues such as digital library copyright protection.
In general, there are several basic requirements for digital watermarking: (1) demonstrate; (2) ineffective; (3) Robust. If there is no robustness (Robustness), the treatment of watermarking and information camouflage technology is essentially identical. In most cases, the information we want to add is imperceptible - some of the specific occasions that use visible digital watermarks, copyright protection markers do not require hidden - and hope that attackers do not destroy the quality of the data itself The watermark cannot be removed. The development of digital watermarking technology is to provide copyright protection for multimedia data, but people find that digital watermarks have other important applications, as shown below, these requirements have promoted the current learning enthusiasm for the current academic camouflage.
(1) Uninterrupted communication means required by military and intelligent institutions;
(2) The national security department's attention to the communication technology of hidden data;
(3) The criminals are transmitted for special information on using regular communication equipment;
(4) The government department screens and controls the information on the Internet's anonymous and encrypted processing, etc .;
(5) Data detection and network verification in e-commerce;
(6) Anti-counterfeiting and data authentication of email, etc.
At present, the enthusiasm for digital watermarking technology is still high, which can also be seen from the academic circles. The enthusiasm of the academia is reflected in the watermarking in the watermark, and the meeting on digital watermarks and data hidden growth has grown very quickly. The enthusiasm of the industry's watermarks is expressed in their significant increase in the level of funding for watermarking. In addition to the research of university and research institutions on watermarking, some international standards have also planned to develop practical digital watermarking algorithms. For example, the Talisman of Europe is to establish a copyright protection mechanism for large-scale commercial infringement and piracy within Europe. Taliman hopes to improve the protection means for video products to increase logo and watermarking. Octalis is a follow-up project of Taliman and Okapi, which is the main purpose to integrate conditional access mechanisms and copyright protection mechanisms.
International Standards Organization also is interested in digital watermarking technology. The upcoming digital video compression standard MPEG-4 (ISO / IEC 14496) provides a frame to allow combination of simple encryption methods and watermark embedding methods. DVD Industry Standards will use watermarking technology to provide replication control and replication protection mechanisms, such as "copy once" or "not allowed", and more.
Although in the past five years, various watermark algorithms are constantly emerging, but digital watermarking technology is still an unproved research area, and there are many problems need to be solved. Theoretical foundation is still very weak. Most water printers Still empiric. It is necessary to work with the following aspects:
(1) Designing a fair comparison and evaluation method for the watermark system, some scholars have some initial research in this regard; but lack of universality and principles, the fragility of the watermark system cannot be fully tested;
(2) From a realistic point of view, the watermark system must reach a balance between the algorithm's robustness, the embedding information of the watermark, and the irreperible sexuality, which involves the principle of robustness algorithm, the construction of watermarking construction model , Theoretical estimation of watermark energy and capacity, the theoretical research of watermark embedding algorithm and detection algorithm. How to determine the balance point is still a problem, and most of the water printers use experience instead of solving this problem from theory;
(3) How to combine watermarking technology with current international image and video compression standards such as JPEG2000 and MPEG-4, and how to apply watermarking technology to DVD industrial standards;
(4) The certificate of ownership has not fully resolved. For many algorithms that have already appeared, the attacker can completely destroy the watermark in the image, or copy a theoretical "original image", which leads to all documents. Veters cannot deliberately provide valid evidence of copyright belonging. Therefore, a good watermark algorithm should be able to provide completely no dispute copyright certificates, and you need a lot of work in this regard. The legal evidence currently uses watermark as copyright protection;
(5) The solution of the audio and video watermarks is not perfect. Most video watermarking algorithms are actually directly applied to the results of their image watermarking and video fields without considering video applications and near real-time features. . From future development, watermarks will have extremely broad application prospects in video and audio sectors, including DVD, other digital products. Therefore, how to design mature, informatively to the international standardized watermark algorithm is still unresolved;
(6) There are many mini, in principle in the original watermark algorithm, but at present, the work at home and abroad has not yet been able to highly refine and in-depth theoretical research on the common problems in different algorithms of the intrinsic links, and thus lack of Digital watermarks have a further study of the theoretical results with guiding significance.
Digital watermarking technology is a frontier research area rise in the international academic circles in recent years. It has a close relationship with information security, information hiding, data encryption, and a key technical issue that needs to be solved in the development of digital libraries. Unlike data hidden is a robust requirement of digital watermarks. Robust can be said to be a very important aspect of watermarking. We have already talked, different application purposes (such as data tracking, data monitoring, and copyright protection, etc.) have produced different design purposes and watermark algorithms. Some watermark methods require raw data to extract embedded watermark information and other no need. In addition, some watermark algorithms that have been released can completely extract watermarks, while others are only checking if the watermark is available. It should be pointed out that these two methods are essentially, and the watermark extraction can be converted to a watermark check, and it is also true. Although the watermark algorithm is often linked to still images, video, and audio, digital watermarks can also be applied to text, three-dimensional grid, and human face. From the current point of view, there is no difference in the inherent watermarking algorithm, and in the embedding process, the PN sequence is modulated and combined with the masking characteristics, and the assumption-based assay based on the test method is used during the test. The design of the watermark algorithm not only considers the general problem of robust problems caused by standard data processing, but also to consider robust problems due to malicious attacks. Several universal attack types have been mentioned previously, and targeted watermark strategies and methods are also in research. In a word, the watermark should have sufficient redundant design to ensure resistance to various attack behavior. In terms of copyright protection, there is also a need to consider some additional aspects. One problem is if there are many watermarks embedded in the image, how to prove that the first first embedded. Solution to this issue may require a method such as a digital time identity or a watermark level. It is also now possible to see that a watermarking algorithm only has sufficient robustness to ensure resolving effective copyright issues, even if the original data can be obtained. Further, the watermark algorithm should be non-symmetrical, unidirectional, irreversible.
Although there is already a commercial watermark system, the research on watermark is still far mature, many problems such as robustness, authenticity differentition, copyright prove, network fast automatic verification, and audio and video watermarks still need relatively perfect solution.
