TABLE OF CONTENTS Deepa Kundur, Karen Su and Dimitrios Hatzinakos
In this chapter, we present the problem of digital video watermarking. Video watermarking is first distinguished from traditional still image watermarking by introducing distinct challenges for video watermarkers and attackers. Common paradigms such as communication theory, steganography, perceptual coding and estimation theory used for algorithm design are introduced to provide a multidisciplinary flavor to the area. State-of-the-art algorithmic research and emerging applications are presented focusing on influential developments in the area. This work is balanced with a discussion of hardware implementation issues and perspectives for digital video watermarking. Trends in video watermarking technology and future directions of research conclude the chapter.
A general digital video watermarking system is illustrated in Figure 1. There are two main processes: watermark embedding and watermark detection. In watermark embedding, a watermark signal usually containing a payload data bit stream U is embedded into an original host digital video sequence X by imposing invisible changes to produce a watermarked video signal X w that it is perceptually identical to X. The embedding may be performed with the use of a secret key K which is a sequence of bits embodying any secret parameters and components of the watermarking system. Detection involves producing an estimate of the watermark payload 
from a possibly tampered version of the watermarked video signal 
. We distinguish the signal used for watermark estimation from X w because of the presence...
