If watermarks could talk: On telltale watermarks and their applications for digital content authentication

The increasing popularity of digital media is a mixed blessing. As opposed to analog media, replication usually produces perfect copies of the original digital content. Digital watermarking has been proposed as a method for embedding a known piece of digital data within another piece of digital data. Many digital watermarking algorithms have been proposed and tested by several parties. This has led to the emerging of new attacks against these algorithms. A good watermark must survive all known attacks to a certain acceptable degree based on the application. In this work, we first present a brief review of the digital watermarking literature. We look into the class of fragile and semi-fragile digital watermarks used specifically for embedding copyright and authentication codes into digital media. We discuss attacks against semi-fragile watermarks, and propose methods to counter these attacks. We also implement a selected semi-fragile watermarking algorithm in order to analyze its performance. We propose some improvements in order to achieve better performance. The performance of the proposed techniques is verified and evaluated by simulations' results. We conclude that without well designed watermarks that can withstand various types of attacks, watermarking may be rendered useless as a scheme for proof of ownership or for discouraging illicit copying of copyright material