Content-dependent watermarking approach based on wavelet tree structure and genetic algorithm
Yi Zhou · Proceedings of SPIE, the International Society for Optical Engineering/Proceedings of SPIE · 2007
In this paper, an approach for image watermarking based on wavelet multi-resolution decomposition is proposed, which presents a new method of designing quantization pedometers based on wavelet tree structure and image content features. According to the new method, select quantization pedometers adaptively to embed the watermarking into the lowest frequency subband of a wavelet domain image. Meanwhile, the performance of the algorithm is enhanced by using genetic algorithm (GA) to seek relevant parameters and by using GA to ensure the optimal balance between the robustness and invisibility of watermarking. Experimental results show that this algorithm produces a satisfying image quality and delivers a stable performance in image processing.