一种基于特征点和伪Zernike矩的数字图像水印算法

A robust image watermark based on feature extraction and pseudo Zernike moments

利用墨西哥帽小波尺度交互的方法提取了图像的特征点,并以特征点为圆心的圆盘中选取部分伪Zernike矩嵌入水印.仿真试验证明,这种基于图像特征点和伪Zernike矩的第二代水印算法.可以抵御剪切攻击和旋转攻击,对频域操作也具有一定鲁棒性.

A robust digital image watermarking scheme that combines image feature extraction and Pseudo Zernike moments is proposed. The goal is to resist both signal processing attacks and geometric distortion attacks such as crop and rotation. We adopt a feature extraction method called Mexican Hat wavelet scale interaction. Several copies of watermark sequence are embedded in the disks which are centered at the extracted feature points. Simulation results show that the wartermark is robust against many image proceses.