Deliberate Distortion of Color Image and Video Resources for Copyright Protection Arash Abadpour 1 and Shohreh Kasaei 2 1 Mathematics Science Department, Sharif Univ. of Tech., Tehran, Iran, email: abadpour@math.sharif.edu 2 Computer Engineering Department, Sharif Univ. of Tech., Tehran, Iran, email: skasaei@sharif.edu Abstract— The easiness of data–flow in digital media, seriously claims the ownership of intellectual material. To solve this problem, researchers have worked on different watermarking methods to embed ownership data into original signals. Unfortu- nately, none of the available techniques have been able to assure an acceptable level of security. Though, recently some senior members of the image processing community are expressing essential doubt about the appropriateness of watermarking for data protection and declare it more as a nonsecure tool for data embedding. In this paper, we propose a fast method for copy protection of color visual objects. The proposed method, deliberately embeds fake edges and color alterations into a given a color image in the way that it carries the content of the original image. Using this damaged image prevents the unauthorized users to pirate the demo version of images and video placed at homepages to help consumers select the images they wish to purchase. The main contribution of the proposed method is that there is a small key for the distorted image that reverts all degradations in a lossless manner. To the knowledge of the authors, this is the first time this kind of copy protection is addressed. While there are literally infinite ways to encrypt a given image, it is proved both experimentally and mathematically that there is no practical chance to crack the code. I. I NTRODUCTION Although, establishment of digital media has been beneficial for enhancing data–flow in the human community, it has resulted in some major drawbacks. In this way, the easiness of storage and doubling of data in digital frameworks, is essentially claiming the copyright integrity of intellectual material. This is the main motivation for researchers to in- vestigate efficient ways to save the ownership of multimedia streams. [1], [2]. One of the main branches of multimedia, is the pictorial data in form of still and moving pictures. Recently, in April 2005, the giants of motion pictures started a serious movements towards banning peer-to–peer sharing of video on the internet. While trials like this have been at most partially successful in providing a safe atmosphere for copy protection, researchers are getting more and more aware to give an effective solution. For example, in watermarking, a legal data is added to the original data to help identifying the original owner. Histori- cally, during last decades different approaches for image (e.g., see [3], [4], [5], [6]) and video (e.g., see [7], [8], [9], [10], [11]) watermarking have been developed. Though, none of them is claimed to be ultimately resistent to all kinds of attacks. Note that the meaningfulness of the “attack–resistance”, reported for many of the available approaches, should be carefully regarded (for a full analysis of attack–resistance see [12], [13], [14] and StirMark [15]). In fact, development of cheating methods and the range of expert people taking part on them, makes it undoubtable that for any watermarking method, there is generally affordable methods of attack to produce non– watermarked images or even embed another watermark [16]. Recently, the integrity of the watermarking methodology and its properness for security is essentially criticized [17]. In the afterwards comments on Herley’s controversial note “Why Watermarking is Nonsense” [17], researchers emphasized that this young field of signal processing is “oversold” and that no method has yet been able to claim “the ability to protect from all possible future attacks” [18], [19]. Researchers em- phasize on nonsecurity–oriented application of watermarking and count on the new methods to come to help in the security field [19]. By the way, M. Barni makes the point clear “Why should we hide the information within the data, when we could more easily use headers or other means to reach the same goal?” [20]. In our opinion this forum is still open. See [21] for a survey of security scenarios in music watermarking and their failure. After all, it seems that the classic methods of data protection are still the best choices. For example, see [22] for a set of guidelines for ultimate image copy protection from a graphics specialist in NASA, containing quotes like “never expose an image in its large size”, and suggesting to use “visual watermarks” and “programming shield”. In a much different approach, a few researchers work on direct copy detection [23], but it is not yet an stable practical tool, too. In this paper, we address the problem of protecting the ownership of a visual object which should exposed to the public. As an application, assume the agencies selling images in their webpages (e.g., webshots.com). the images should be presented to possible consumers, but the viewers should not be able to save the images for their own and then probably use them in professional publication. Although, there are some programming shields for this purpose, we believe they all will be soon cracked, as similar guards did. The classic solution to this problem is to put the images in a restricted manner, for examples, cropped, badly compressed, or down–sampled. In this way, for selling image I , the agency should reserve room for two images. While the original image, I , should be saved in a high–security zone, system should produce and expose ˜ I , the restricted version of I , to the public. In this strategy, a overall redundancy of ‖ ˜ I ‖/‖I ‖ is forced to the system. where, 369 0-7803-9314-7/05/$20.00©2005 IEEE 2005 IEEE International Symposium on Signal Processing and Information Technology