一种基于DWT和HVS的图像版权保护研究  

ON DWT AND HVS-BASED IMAGE COPYRIGHT PROTECTION

作　　者：谢泽会[1] 曹卫锋[1] 

机构地区：[1]郑州轻工业学院电气与信息工程学院,河南郑州450002

出　　处：《计算机应用与软件》2010年第8期75-77,共3页Computer Applications and Software

基　　金：国家自然科学基金资助(60773122)

摘　　要：提出一种基于离散小波变换(DWT)的数字水印进行图像版权保护的新方法。将一幅有意义的二值图像作为水印来隐藏,先将水印图像使用推广Arnold的变换进行置乱后和图像同时进行多尺度分解,然后将分解后的水印系数根据人类视觉系统(HVS)特性自适应地嵌入到具有相同尺寸的低中频系数中,重构得到水印图像。实验结果表明,该算法具有较好的不可见性、鲁棒性和安全性。A new algorithm of digital watermarking based on discrete wavelet transform（DWT） is presented for image copyright protection. It conceals a meaningful binary image as the watermarking, and scrambles the watermarking image by generalized Arnold transform first, then makes multi-scale decomposition on it together with the image at the same time. After that, the decomposed watermarking parameters are adap- tively embedded into the low and medium-frequency sub-band with same size according to the feature of the human visual system（ HVS）, and the associated coefficients are restructured to obtain the watermark image. Experimental results show that the algorithm has fairly good imper- ceptibility, robustness and safety.

关 键 词：离散小波变换 数字水印 人类视觉系统 鲁棒性 重构 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]