Digital media application, multimedia security, copyright protection are becoming important in today’s world as its distribution over network is increasing rapidly, indeed left a concern for the owner about their ownership protection. For the copyright protection of the digital applications, digital watermarking is introduced that enables to mark the digital image with undetectable secret information without affecting its perceptual quality known as watermark to cover the various digital data security such as copyright protection, access control and authentication for still images , video, multimedia products. This thesis presents a novel technique by using principal component analysis and discrete wavelet transform resulting into high imperceptibility where no noticeable difference is found between the watermarked video frame and original frame thereby increasing the performance of watermarking algorithm. The concept of key-dependent basis function is introduced, its applications are discussed to secure robust watermarking for copyright protection and to design a secure public black box watermark detectors. Thereby, it overcomes the possible security weakness of non-adaptive as well as global schemes that execute watermark patterns with a small number of publically known basis function. Projections of an image is embedded within the watermark onto the secret set of key-dependent functions (patterns). The robustness of the watermarking scheme with respect to filtering, loss compression and combinations of many other attacks is studied. Conclusively, we propose a candidate for a watermarking scheme that enables the construction of a secure public watermark detector.
Index Terms: Digital Watermarking, Principal Component Analysis, Discrete Wavelet Transform, Copyright Protection
A remarkable growth is seen in the use of digital video now a days including various applications which are involved in our day to day life like video-on-demand, digital television, video-conferencing, advertising, surveillance, entertainment and distance learning. Many user experience digital video when they watch a motion picture recorded on a digital videodisc (DVD) or downloaded over the internet. The digital video’s proliferation are encouraged into many more applications by improving the compression technology which results into better authoring and editing tools as well as more available bandwidth in digital communication networks and lower cost capture and display devices.
Digital representation offers many advantages for processing as well as distributing video and other types of information world wide. First, digital software programs offer unprecedented flexibility in presenting, creating manipulating and editing the digital information. Analog devices lack the flexibility, malleability and extensibility of processing of the software. Second, digital data is allowed to be distributed and disseminated on a wide scale through digital communication network like internet. On some of these networks, existing open and proprietary protocols such as the World Wide Web allow any user to easily and inexpensively exchange, provide, find and obtain the digital information. Lastly digital information can be processed, and in particular, copied without introducing loss, degradation, or noise. For example, an unlimited number of perfect copies can be produced from a single digital video signal. In contrast, the addition of noise into a copy from analog signal processing is unavoidable.
A watermark is a form, image or text that is impressed onto the paper, which provides the authenticity and the ownership. Digital watermarking is an extension of this concept in the digital world. A phenomenal and rapid use of the internet has highlighted the need for developing a mechanism to protect ownership of digital media. While the aforementioned advantages offers immense opportunities for creators, the ability to make perfect copies and the ease by which those copies can be distributed and also facilitate misuse, illegal copying and distribution that is piracy, plagiarism and misappropriation. Content owners and creators are concerned more about the consequences of the illegal copying and distribution on the massive scale. This problem is not merely theoretical as illegal copying and distribution of copyrighted materials has given an arise to the economic damage which is estimated to be in the billions of dollars. Recently, popular internet software based on peer to peer architecture such as Kazaa, BitTorrent, eDonkey, and Gnutella have been used to share or distribute the copyrighted movies, software, music and other materials. New peer-to-peer systems may encrypt the data being preserve the anonymity of its users, shared, support a larger number of users and be more robust. These advances in peer-to-peer systems will create considerable challenges for copyright enforcement. Thus, there is a great desire for methods which can preserve the economic value of digital video and protect the rights of content owners.
1.2 SCOPE OF WORK
The reproduction, manipulation and the distribution of digital multimedia ( images, audio and video) via networks becomes faster and easier. As the proprietors and creators of the digital products are aware of illegal copying of their products therefore, security copyright protections are important issues pertaining to multimedia applications and services.
Earlier, the watermarking techniques were proposed for these aforesaid purposes in which the copyright information is embedded into multimedia data for protecting the ownership. Consequently, research is now being focused on watermarking schemes to protect multimedia information. The most suitable technology that can serve this purpose is none other than digital watermarking. Multifarious watermarking schemes have been proposed to camouflage copyright marks and other information in digital images, audio, video and other multimedia objects.
One of the biggest technological events of the last two decades was the invasion of digital media in the entire range of everyday life aspects. Computer techniques can easily manipulate and store the digital data efficiently and effectively that too with a high quality. Furthermore, digital data can be transmitted in a fast and inexpensive way through data communication networks without losing quality. A watermarking is adding ‘ownership’ information in multimedia contents to prove the authenticity along with embedding a data which is an unperceivable digital code, namely the watermark that carries information about the copyright status of the work to be protected. Continuous efforts are being made to device an effective watermarking scheme but techniques proposed so far do not seem to be robust to all possible attacks and multimedia data processing operations. The abrupt increase in watermarking interest is most likely due to the increase in concern over IPR. Generally, the watermarking of still images, audio and video demonstrate certain common fundamental concepts. Innumerable watermarking applications reported in the literature depend upon the services we wish to apply. Thus, watermarking techniques may be relevant in numerous areas which may include copy protection, copyright protection, fingerprinting, and temper detection etc. . The manner in which information is embedded, watermarking schemes can be classified either as spatial domain (the watermarking system directly changes the main data elements, wiz pixels in an image to hide the watermark data) or transformed domain (the watermarking scheme alters the frequency transforms of data elements to hide the watermark data). The last technique has proved to be more robust than the special domain watermarking .
Several reversible transforms are used to transfer an image to its frequency representation like discrete cosine transform(DCT), discrete fourier transform(DFT), or discrete wavelet transform(DWT), . Even though spatial domain based techniques cannot sustain most of the common attacks like compression, low-pass or high-pass filtering et-al, researchers present spatial based techniques too .
Since, monetary impact of some of the application areas like copyright protection, fingerprinting are very high and till now no successful algorithm seem to be available to prevent illegal copying of the multimedia contents, the ultimate goal of this thesis work is chosen to develop watermarking schemes for images (which are stored in spatial domain as well as transformed domain) which can sustain the known attacks and various image manipulation operations.
1.3 RESEARCH GOALS
As video copyright protection is the main issue of the thesis, a robust video watermarking scheme is required and to achieve this, three parameters are to be taken into consideration that are as follows:
1.3.1 SECURITY IN MULTIMEDIA COMMUNICATION
In a decade years ago, multimedia documents are rarely available to the mass consumer market. Although, as the rapid development of pervasive digital information technology is seen, everyone have dense and portable storage, good quality of video compression techniques, high processing power and increased network benefit in their personal computer and this may lead to generation of another crisis as in the users have ability to tamper, generate copies, redistribute illegally the digital content. To take off in an e-commerce setting, the security issues are to be resolved regarding digital multimedia products and services.
According to research findings, a unique solution is noticed which integrates the secure information into the content of the video frames which must be inseparable and that particular information must be invisible and should not get affected by any attacks and, therefore that unique solution is the digital watermarking which strive to achieve the goal by embedding a retrievable watermark directly into the video frames.
1.3.2 CONTENT PROTECTION
Content protection is a challenging problem which involves conflicting interests. Content owners wish to ensure that their intellectual property is not misused, illegally copied or distributed. Although, device manufacturers have in mind to keep their products simple and inexpensive, but implementation of technological content protection leads to measurable increase in the cost and complexity of devices.
Potential customers may shun such devices if the device manufacturer does not take into consideration to minimize the risk of introducing devices to market which restrict or limit the activities of their users. Users desire their legal privileges (such as first sale and first use) and privacy to be safeguard. Users do not wish to pay for devices which are more expensive, less compatible (interoperable), and more restrictive of their activities. Though the primary beneficiaries for such systems are the content owners and providers then also users are cautious about the burden of the cost of content security system. While, it is not clear where the balancing point amongst the diverse interests ultimately, technical and legal initiatives lies towards content protection, the Digital Millennium Copyright Act (DCMA) was the first legislation in a series of attempt by the U.S. Congress to update U.S. copyright law for the digital era. President Clinton signed the act into law on October 28, 1998. The DMCA prohibit circumvention of technological measures used by copyright owners to European works. Similar provisions appear in their defense Copyright Directive (EUCD), or Directive 2001/29/EC of the European parliament. The EUCD obliges the Member States to call into being legal protection against the circumvention of technical security measures as well as against construction equipment for sale or trade, offering the primary object of which is to circumvent these technical security measures. EU member states are currently implementing the directive as per desired be it with a much slower pace than what the implementation deadline of December 2002 called for. In addition, some national parliaments have rejected proposals to implement the EUCD. To complement the legal initiatives that were taken, technical measures are sought by content owners to protect their creations. Generally technical content protection measures use three approaches: access control, copy protection and material tracking.
The objective of access control is to ensure that video is accessible only under the rules or conditions specified by the content owners. For example, the owner may specify that only certain users may view the video, or payment is required each time when the video is viewed, or that the video can be viewed for limited number of times. Access control alone offers limited protection, however, because at some point the video will be provider to user. If this occurs then access control will not be able to prevent the user from creating illicit copies or misusing the video. Copy protection prevents or hinders the creation of copies or misusing the videos. Copy protection prevents or hinders the creation of copies. Copies of a video can be digital created, it is displayed as video recording or to the user.
Recording the video as it is displayed for the intent of circumventing or removing technical content protection measures is known as exploiting the ‘analog hole’. Copy protection system identifies protected video and then uses some means to prevent or make more difficult to creation of a copy. Unfortunately, copy protection is very challenging and techniques for copy protection have consistently been defeated. Content tracking is a means to protect video even if access control and copy protection are circumvented. The content monitor, each legitimate copy of the video is personalized or individualized by embedding information indicating the user who has custody of that particular copy. If in a suspicious location (such as being shared using P2P software), a copy of the video is discovered, an investigator may remove embedded information from the copy and determine source or origin of the suspect copy. The content owner may decide to initiate action if many suspect copies are found.
Access control has often been addressed with help of encryption technique. The process of scrambling of data into an unintelligible form is known as encryption. The original data is known as the plaintext and the scrambled data is known as the cipher text. The inverse process of obtaining the plaintext from the cipher text is known as decryption. Encryption provides confidentiality because a secret key is necessary for decryption. Traditionally, encryption has been used to ensure the confidentiality of sensitive information transmitted through an insecure communication channel. For access control, video is encrypted and the decryption key is provided only after the access conditions have been satisfied. Obtaining the encrypted video alone does not allow the video to be displayed. The significant limitation of encryption is that it offers no protection once the video is decrypted. This alone implies that encryption is not sufficient for content protection and another method is needed to protect the video after decryption. Watermarking has been proposed to provide more lasting protection.
Following contributions have been carried out throughout the research work:
‘ Till now there is no ‘Generic’ nature in the watermarking algorithms available. More precisely, if certain approach is applicable for a gray level image, the same approach does not work for the other formats of an image.
‘ Even if image watermarking algorithms for gray color are extended for RGB color images, the maximum work has been done for (y-luminance) color channel only because human eyes are less sensitive to detect the changes in (y-luminance) color channel.
Attack-Impact free analysis has been carried out for color channel that may be affected by a particular attack . Therefore, apart from choosing digital video watermarking as a major problem, we have chosen to identify the suitability of a color channel with respect to attack for multi-color channel images. We also decided to explore the ways like attack impacts may be minimized before the watermark embedding process.
‘ Predominately in research papers, once the watermarking scheme is acknowledged, it is applied to all test images. As each image is different and has its peculiar characteristics and after embedding the watermark data by a particular watermarking scheme, its performance against a particular attack may not be similar with other image. No study is conducted to make the embedding scheme based on some image characteristics.
Therefore, we have resolved to establish the relationship between the performance of watermarking scheme and the cover image characteristics.
‘ Mostly watermarking schemes are developed in a way that first a scheme is developed based on the extension presented one and then check its performance against the common image manipulations and known attacks. There are huge financial implications of watermarking schemes, but no scheme has been developed, which is, by design, resistant to at least one attack, to ensure that, a particular attack having most financial issues cannot be conducted by an attacker. Therefore, we decided to design watermarking schemes such that an inherent nature which can be embedded to guarantee that at least one serious attack having most financial implications cannot be conducted on watermarked images.
1.5. ORGANIZATION OF THESIS