Embodying Intelligence in Multimedia Data Hiding

Abstract

Data hiding is the process of hiding an amount of data, called secret message or watermark, into a cover media that may be audio, video, or image signal, in an imperceptible way for different purposes. Due to the advances in generation, storage, and communication technology of digital media, the necessity and importance of data hiding has drawn the attention of many researchers all over the world resulting in a lot of variants of data hiding scheme. This chapter provides a detailed review of the basic concepts of data hiding and a survey of its major variants developed so far. The chapter classifies the data hiding schemes based on several aspects of data hiding comprising artificial intelligence. Furthermore, the chapter also provides the recommendations to the interested users for further research. This review may be a useful insight and a good source for the users who are interested in the application of artificial intelligence in data hiding.

Abstract

Databases play a pivotal role in all domains of technology, encompassing data mining, medical records, stock market data, e-commerce etc. With this elevated need for databases and their wide distribution in the web sphere, their security has become a major concern today. It is in this context that watermarked protection of databases has started receiving increasing attention from researchers. We present two of the most important security concerns related to protection of relational database: (i) Ownership proof and (ii) Tamper detection. We explain the functions of various components of a Robust Watermarking Model (RWM) to resolve ownership issues and a Fragile Watermarking Model (FWM) to deal with Integrity issues. We discuss various attacks that are possible on databases and present the appropriate security solutions to overcome them.

Abstract

Digital image watermarking is the process of concealing secret information in a digital image for protecting its rightful ownership. Most of the existing block based singular value decomposition (SVD) digital watermarking schemes are not robust to geometric distortions, even if for some special distortions, such as multiples of 90° rotation of integers and image flipping, which change the location of pixels but have no effect on the value of the image. Therefore, to overcome the problems mentioned here, this chapter proposes a novel image watermarking scheme by redistributing the image and applying some normalization operators. Subsequently, this image is segmented into non-overlapping blocks and SVD is applied to each block to get the largest singular value. The watermark is scrambled by using the Arnold cat map and then the bits are embedded in the blocks by quantizing the largest singular value of each block. In addition, the firefly algorithm (FA) is applied to obtain the quantization step (QS) optimally to improve the fidelity and the perceptual quality of the watermarked image. To investigate the robustness of the scheme several attacks are applied to seriously distort the watermarked image. Empirical analysis of the results has demonstrated the efficiency of the proposed scheme.

Abstract

The growth of multimedia technology has lead to easier and faster ways of communication. At the same time, it has broadened up the gap between security and transmission of various multimedia objects. Video and image are the most popular shared objects over the open network through various portable devices. Digital watermarking is the data hiding technique where the owner’s content is inserted into the original content (i.e. image and video) for ownership verification. The process of embedding the watermark in real time watermark has more significance because there is no delay between the capturing of image/video and the insertion of the watermark. The hardware implementation is an ideal choice of real time watermarking since it outperforms the software implementation in terms of area, power and speed.
In this chapter, two different algorithms are proposed, one is based on a variable block size concept and the other uses AC prediction technique. The performance of the algorithms is calculated and compared with similar previous work. The algorithms are prototyped on the FPGA to validate the real time performance of hardware.

Abstract

The development of photocopy technology and manufacturing technology provides many possible ways for adversaries to fake and modify paper-based legal documents such as certificates, licenses, and invoices without being detected. It makes managing the paper-based legal documents a challenging task. Hence, managing and authenticating digital legal documents becomes a practical and requisite task in daily life.
To relieve the efforts of document management and authentication, in this chapter, we propose two security verification systems for digital media: the public key encryption based system and the Chinese Remainder Theorem (CRT) based system. Both systems are able to produce and verify secure electronic documents while keeping the owners’ legal privacies secret. Throughout the chapter, we present and analyze the background, structure, security performance, and implementation of the two schemes.

Abstract

Recent advances in genetic engineering have allowed the insertion of artificial DNA strands into the living cells of organisms. These advances made it possible to insert information into a DNA sequence for the purpose of data storage, watermarking, or communication of secret messages by using substitution ciphers and other methods.
While many algorithms have been developed to hide and insert messages into DNA sequences, there are only few approaches of discovering such messages. The ability to detect, extract, and decode messages from DNA is important for forensic data collection and for data security. One of our goals is the development of a software toolkit that employs a combination of several algorithms to decode a message written in DNA symbols. Genetic Algorithms (GA) have traditionally not been successful in solving substitution ciphers, but we have developed a new approach that uses a GA in combination with the Wisdom of Artificial Crowds post processing Algorithm to overcome these limitations. Our approach is successful in breaking substitution ciphers that encode messages in DNA. Results show that this approach delivers significantly more accurate results than a dictionary approach or a GA by itself.