利用人眼对比敏感度视觉特性和图像的纹理特征，提出了一种彩色图像的水印算法。算法首先依据图像的纹理特征，将图像进行分割；再将水印嵌入到图像纹理复杂的部分，并利用人眼对比敏感阈值控制水印的嵌入量，从而实现水印的嵌入；最后利用逆过程实现水印的提取。同时对3幅不同复杂纹理的图像进行仿真实验，并对3幅含水印图进行5种方式的攻击测试，且与近些年国内外报道的8种水印算法进行对比分析。结果表明，在中间剪切1/4区域、质量因子为20%的压缩和旋转90°较强的攻击下，提取水印的NC值和含水印图的SSIM值均仍能达到0.8以上，且比8种水印算法具有更好的鲁棒性。综合结果表明，在水印嵌入过程中，提出的算法能够有效解决水印容量与视觉透明性、鲁棒性之间的平衡，是一种行之有效的图像水印算法。

Using the contrast sensitivity characteristics of human visual system (HVS) and the texture features of image, a watermarking algorithm for color images is proposed. In the algorithm, the image is firstly divided according to the texture features of image, then the watermark is embedded into the complex texture part of image, and the embedding amount of watermark is controlled by the contrast sensitivity threshold of HVS, which finish embedding the watermark. Finally, the watermark is extracted by the inverse process. Further, three images with different complex texture are carried out simulation, and three watermarked images are tested by 5 types of attacks. And this watermarking algorithm is compared with 8 ones reported in recent years. The results show that under the stronger attack, namely the intermediate 1/4 area of image is sheared, the compression quality factor is 20% and the image is rotated by 90°, the NC values of extracted watermarks and the SSIM values of image embedded watermark can still reach over 0.8, and it has better robustness than the reported 8 watermarking algorithms. The comprehensive results show that the proposed algorithm can effectively solve the balance between watermark capacity and visual transparency, robustness, which is an effective image watermarking algorithm.