Article Preview
Top1. Introduction
Recent advances of digital technologies make alteration of multimedia data simpler and undetectable by human audible-visual system (Darko & Borko, 2005). One of the primary reasons could be digital contents are easily accessible via Internet and can be replicated and redistributed by illegal means. Use of digital content with proper acknowledgement for academic purpose or non-profit business is a common practice. But, forging data, usage of someone data without due citation, tampering data and blackmailing someone credential, making money from stolen data, etc are major concerns in multimedia security. When a bad person is getting away by doing these practices without being caught then there are two important issues arise - on one hand, the bad person knows that s/he can enjoy her/his life in this way without expecting any punishments; and on the other hand, it is because of poor content protection and detection mechanism by the content owner. In other words, preventing bad persons from fabricating content and detecting both fabricated data and creator are challenging tasks in securing multimedia data. In order to protect multimedia data, some techniques such as digital watermarking, scrambling, authentication, integrity or a combination of these play an important role. This paper focuses one of the techniques, namely digital watermarking with cryptographic primitives.
Digital watermarking (Darko & Borko, 2005). is a technique that inserts a piece of information into a target image, which can be later extracted for a variety of purposes and/or detected in case of malpractices. A watermark can be a binary string, a logo, or some intended features of the multimedia content. Whatever maybe the watermark content, the watermarking technique needs to be designed in such a way that an unauthorized entity should not be able to identify any information about the watermarked content and position of it. Whereas, the content owner or authorized entity should be able to identify the watermarked information from the image as and when needed.
Watermarking techniques (Darko & Borko, 2005; Mandhani, 2003; Hwang, Chang, & Hwang, 1999) can be broadly classified in two categories: visible watermarking and invisible watermarking. Visible watermarking is the classical way of watermark generation, extraction and deletion, where the watermark is inserted into a cover object. Although visible watermarking is useful for identification purpose, it is not secure watermarking for applications such as image authentication, image alteration and copy protection. In contrast, invisible watermarking is the modern approach for digital image authentication and piracy detection. A typical process flow of embedding and extraction of digital watermark in an image is shown in Figure 1.
Figure 1. Embedding and extraction of watermark