Region of Non-Interest Based Digital Image Watermarking Using Neural Networks

Copyrights protection of digital data become inevitable in current world. Digital watermarks have been recently proposed as secured scheme for copyright protection, authentication, source tracking, and broadcast monitoring of video, audio, text data and digital images. In this paper a method to embed a watermark in region of non-interest (RONI) and a method for adaptive calculation of strength factor using neural network are proposed. The embedding and extraction processes are carried out in the transform domain by using Discrete Wavelet Transform (DWT). Finally, the algorithm robustness is tested against noise addition attacks and geometric distortion attacks. The results authenticate that the proposed watermarking algorithm does not degrade the quality of cover image as the watermark is inserted only in region of non-interest and is resistive to attacks.