基于彩色图像Tetrolet变换的频数质心零水印算法

针对传统数字水印算法的透明性随嵌入强度的增加而下降的问题, 已有许多零水印算法提出。然而大多数算法或是鲁棒性和稳定性较差, 或是基于灰度图像设计的。本文提出一种基于彩色图像Tetrolet变换的频数质心零水印算法。该算法先提取彩色载体图像YCbCr空间的亮度分量, 再对亮度分量进行Tetrolet变换获得其低频系数, 再通过分块非负矩阵分解, 采用计算频数质心的方法构造特征矩阵。对水印信息依次进行Polar码编码, 二维交织构造出交织矩阵; 最后将含有水印信息的交织方阵和特征矩阵进行异或操作生成零水印。仿真结果证明, 该零水印算法可以有效抵抗非几何攻击、几何攻击以及组合攻击, 水印提取归一化相关系数值在0.9以上, 具有较强的鲁棒性和稳定性, 且具有良好的实用性。

Frequency centroid zero watermarking algorithm based on Tetrolet transform of color image

In order to solve the problem that the transparency of traditional digital watermarking algorithms decreases with the increase of embedding strength, many zero watermarking algorithms have been proposed. However, most algorithms are not robust and stable, or are designed based on grayscale images. In this paper, a frequency centroid zero watermarking algorithm based on color image Tetrolet transform is proposed. The algorithm first extracts the luminance component of the color carrier image in the YCbCr space, then carries out the Tetrolet transform to obtain its low-frequency coefficients. Then through the block non-negative matrix decomposition, it uses the method of calculating the frequency centroid to construct the feature matrix. The watermark information is encoded by Polar code in turn, and the interleaving matrix is constructed by two-dimensional interleaving. Finally, the zero watermark is generated by exclusive OR (XOR) operation between the interleaved square matrix and the feature matrix containing watermark information. Simulation results show that the zero watermarking algorithm can effectively resist non-geometric attacks, geometric attacks and combination attacks, and the watermark extraction normalized coeffecient (NC) value is more than 0.9, with strong robustness, stability, and good practicability.