【作者】 陈宇； 曲芳； 胡建斌； 陈钟；

【Author】 Chen Yu,Qu Fang,Hu Jianbin,and Chen Zhong (Institute of Software,School of Electronics Engineering and Computer Science,Peking University,Key Laboratory of High Confidence Software Technologies(Peking University),Ministry of Education,Beijing 100871)

【机构】 北京大学信息科学技术学院高可信软件技术教育部重点实验室；

【摘要】 鲁棒性是数字水印极为重要的指标.目前大多数文献中的鲁棒性水印方案对常规信号处理攻击均具有良好的鲁棒性,但抗几何攻击的鲁棒性相对较弱.提出了一种新的灰度级鲁棒性水印算法,对常规信号处理攻击和几何攻击均具有良好的鲁棒性.水印算法的鲁棒性由以下3方面获得:1)选择载体图像的内切圆盘作为Zernike矩的计算域,选择内切圆盘的内接正方形作为水印的嵌入区域.2)在载体图像嵌入区域的层叠DWT-DCT域嵌入水印,获得对常规信号处理攻击的强鲁棒性.3)保存原始含水印图像的2个Zernike矩作为校正几何攻击的参数,获得抗几何攻击能力.实验结果表明,该水印算法嵌入容量大,不仅对大角度旋转和大尺度缩放等几何攻击具有很好的鲁棒性,而且可以有效地抵抗有损压缩和常规信号处理攻击.更多还原

【Abstract】 Robustness is of great importance for digital watermarking.Most watermarking algorithms proposed mainly address the robustness to common signal processing while neglecting geometric attacks.In this paper a novel gray level watermarking algorithm is proposed,which can provide robustness to both geometric attacks and common signal processing.The robustness is achieved from three aspects.1)Choosing the inscribed disk of the host image as the Zernike moments computation domain and the square in the disk as the watermarking embedding domain.2) Embedding the watermark in the cascade DWT-DCT domain in order to get robustness to common signal processing. 3)Saving two selected Zernike moments of the original watermarked image to detect and correct possible geometric attacks.The experimental results demonstrate that the proposed algorithm is robust to a wide range of rotation attacks and scaling attacks,and as well,a variety of other distortions such as lossy compression,common signal processing.更多还原