基于模板的几何攻击估计及图像水印算法研究

The estimation of geometry attack based on the template and the research on image watermarking algorithm

首先提出了一种简单的方框"模板",并估计了该模板抵抗各种几何攻击的能力,在此基础上,一种小波域自适应的抗几何攻击水印算法被提出,该算法充分利用人眼视觉系统掩蔽特性,在DWT域的低频子带中完成水印信息的嵌入工作,并利用所嵌入模板来估计和确定图像所经历的几何变换攻击.实验结果表明,所提出的算法不仅对常规的图像处理具有较强的鲁棒性,而且能够抵抗各类几何攻击,如平移、镜像、旋转、缩放和转置等,此外,该算法具有较好的不可感知性.

This thesis first proposes a simple pane＂template＂,makes estimation for template on resisting all kinds of geometric attacks,and proposes a self-adaptive watermarking algorithm in DWT domain which robust to geometric distortion.In this algorithm,according to the conceal quality of Human Visual System（HVS）,a watermark was embedded in low band of DWT domain adaptively.It makes use of the template to estimate and confirm the geometric transformation of the image.The experimental results show that the algorithm is simple and efficient.The proposed watermarking algorithm is imperceptible and robust against geometric distortions,such as translating,mirroring,rotating,zooming and transposing.At the same time,it is robust to general image attacks.