An Image Watermarking Approach Based on Artificial Fish Swarm Algorithm

In this paper, an artificial fish swarm algorithm was developed to select a proper basis of wavelet packet transform used in the robust image watermark. In the proposed algorithm, watermark is viewed as a vector. To embed the watermark, a vector composed of wavelet coefficients selected from wavelet packet transform of an image will then be modified. The performance measure is the quality of the watermarked images. Both artificial fish swarm and the watermark embedding process will try to optimize the image quality, so the performance will be better than directly embedding the watermark into wavelet domain coefficients. Simulation results obtained from the approach based on the artificial fish swarm algorithm are compared with the heuristic based approach. The experimental results show that the proposed approach in terms of image quality is encouraging when compared to the heuristic approach.