SVD BASED OPTIMIZATION OF FIDELITY WITH NEW GENETIC ALGORITHM

Abstract

Digital watermarking is a process that embeds watermark (data, image, audio and video) in to a host image (data, image, audio and video), such that the watermark can be detected or extracted later to make an authenticity about the object. It is used to resolve the problem of security of data, ownership, and copyright protection. For the purpose of designing and developing a new watermarking algorithm in those application areas, the most important properties are imperceptibility (fidelity) and robustness. But there is a trade-off between robustness and fidelity. This dissertation presents SVD based Centre of mass selection operator (CMGA), to find the optimum locations for the insertion of watermark in a cover image, by focusing on the issue of optimizing fidelity with good robustness. Fidelity is the main issue in different watermarking based on spatial and frequency domain watermarking. In this dissertation we used a Centre of mass selection operator (CMGA) based GA, to investigate the variation of maximum fitness based on the low MSE value of watermarked image, against embedding strength, number of genes, various watermarks. Proposed results show that, this New Genetic Algorithm provides much better solutions than the simple Tournament selection GA.