A Novel Multi-objective Genetic Algorithm Optimization for Blind RGB Color Image Watermarking

Abstract

In this paper, we propose a new Multi-Objective Optimization (MOO) approach using the second version of non-dominated sorting genetic algorithm (NSGA-II) for Blind RGB color image watermarking. The design of an optimal watermarking for a given application always involves a tradeoff between imperceptibility requirement and robustness requirement. Therefore, the image watermarking can be considered as a multi-objective optimization problem. In the process of searching an optimal solution, the trade-off between robustness of the watermarked images against attacks and imperceptibility of the watermarks is presented by Pareto-Front discovered by NSGA-II. Experimental results show improvement both in imperceptibility and robustness under a wide range of attacks.