ABSTRACT

Digital Watermarking offers techniques to hide watermarks into digital content to protect it from illegal copy or reproduction. Existing techniques based on spatial and frequency domain suffer from the problems of low Peak Signal to Noise Ratio (PSNR) of watermark and image quality degradation in varying degree. Earlier technique based on Full Counterpropagation Neural Network (FCNN) used the concept of embedding the watermark into synapses of neural net rather than the cover image to improve PSNR of watermark and to prevent image quality degradation. However, problems like "Proprietary neural net" and "sure win" still exist as explained in this work. This paper is an attempt to uncover and solve these problems. FCNN can be practically employed to obtain a successful watermarking scheme with better time complexity, higher capacity and higher PSNR with the suggested modifications.