Subjective and objective watermark detection using a novel approach: Bar-coded watermarking

Abstract

Many digital watermarking algorithms are proposed in the literature. Broadly these watermarking algorithms can be classified into two main categories. The first category of algorithms uses a pseudo random Gaussian sequence (PRGS) watermark whereas the second category of algorithms uses a binary logo as a watermark. The main advantage of PRGS based watermarking scheme is its ability to detect the presence of watermark without manual intervention. However the main drawback is calculating reliable threshold value. In the similar manner the main advantage of binary logo watermark is that there is no need to calculate threshold value but requires manual intervention to detect the presence of watermark. The advantage and disadvantage of either approach is quite clear hence it would be a good idea to design a watermarking scheme which inherits the advantages from both these approaches. In this paper we present one such approach which is termed as bar-code watermarking. The proposed scheme offers objective as well as subjective detection. A PRGS sequence watermark is represented as a bar-code on a binary logo and embedded in the host image. Watermark detection can be either done subjectively or objectively.