A GA-based Optimal Image Watermarking Technique

This paper presents a kind of GA_based image watermarking algorithm. Considering two conflicting objectives: robustness and invisibility in watermarking system, it is always difficult to satisfy them at same time. We resolve this problem via genetic algorithm (GA). With a proper fitness function in the watermarking system, GA search for the optimal parameters of embedding strength and times of Arnold transform to improve the performance of watermarking algorithm. The simulation results using various watermark attacks show that the proposed method can improve the quality of the watermarked image and give almost the same robustness of the watermark.