轮廓波变换的激光图像水印技术研究

Image watermarking algorithm based on contour segmentation

针对小波变换激光图像水印处理中抗加噪的鲁棒性较差的问题,提出了基于轮廓波变换的激光图像水印算法,首先采用多频段线状特征分解方法尺度分解激光图像,采用多尺度的轮廓边缘特征提取算法进行激光图像的子块区域分割,然后进行水印嵌入,对嵌入水印信息的子块执行反傅里叶变换,得到含水印的图像,最后通过仿真实验进行性能测试,结果表明,该算法的激光图像水印嵌入处理的峰值信噪高,归一化相关系数大,说明算法的抗加噪鲁棒性高,性能优于传统方法。

The image watermarking processing can improve the image encryption and copyright protection. To improve the robustness against noise with wavelet transform color image watermarking, it proposed a color image water-marking algorithm based on Radon wavelet transform. First using multi band linear feature mathod to scale decomposi- ing laser image. Using multi-scale edge feature contour extraction algorithm for laser image block segmentation and then the watermark is embedded by Radon transform. The watermark information embedded on the sub block was implemented inverse Fourier transform to get the image with watermark. Finally, through the simulation performance test, results show that this proposed algorithm for laser image watermarking processing has high peak signal to noise ratio, big normalized correlation coefficient. It has high robustness against noise algorithm and better performance than the conventional method.