Performance comparison of watermarking techniques of various domains

Digital watermarking has gained a lot of importance for digital media recently. It deals with hiding secret bits of information with a digital content as a cover. This is very useful for many applications like copyright control, authenticity of documents, captioning, broadcast monitoring etc. Various techniques for watermarking have been proposed by various researchers in the recent past. Many of these techniques have problems related to robustness, fidelity and payload capacity of watermarks. These techniques also exhibit the problems of trade-off between robustness and fidelity. Techniques based on Backpropagation neural network, Full counterpropagation neural network and Adaptive resonance theory(ART) have been developed and tested by the author for watermarking . The results have shown positive results in increasing robustness, fidelity and payload capacity of watermarks. The brief discussion of various techniques with their results are presented this paper which forms the basis of comparison between the these techniques with some of the conventional techniques whose results have also been shown.