A Novel Blind Attacking Algorithm of Digital Watermarking Based on Sparse Analysis in Wavelet omain
In the paper, a novel effective blind attacking algorithm is proposed, which uses for the full-blown methods of sparse analysis on the condition of Independent Component Analysis (ICA). In the calculation of separating matrix, we make use of two objective functions to enhance the accuracy of the separation matrix: one is to maximize the non-Gaussian of the original image, the other is to minimize the non-Gaussian of the watermark sequence. The simulation results on the one hand verify the validity of the proposed algorithm, on the other hand show that the attacking effect is enhanced to a greater extent under the premise of ensuring the visual quality of watermarked image.
ICA sparse analysis digital watermarking wavelet
Zhang zheng-bao Yang su-min Tan yue-hui CHU chuan-bin
Department of Computer Engineering, Ordnance Engineering College, Shijiazhuang 050003, China
国际会议
第八届国际测试技术研讨会(8th International Symposium on Test and Measurement)
重庆
英文
720-723
2009-08-01（万方平台首次上网日期，不代表论文的发表时间）