基于模糊聚类分析和HVS的自适应彩色图像水印算法

The adaptive watermarking algorithm of color images based on fuzzy clustering and HVS

结合模糊C均值聚类算法和人眼视觉特性,提出了一种新的自适应彩色图像水印算法。首先,将彩色图像经模糊聚类分析,选取出适合于水印嵌入的位置;然后,分别在R、G、B3个通道中利用小波域的视觉掩蔽特性自适应地修改水印嵌入强度,提取时,不需要用到原始图像。实验结果证明,含水印的彩色图像没有出现任何可感知的视觉失真,同时,该算法对一定的图像处理操作具有较强的鲁棒性。

Based on the fuzzy C means clustering algorithm and human visual characteristics, the paper proposes a new adaptive watermarking algorithm of color images. After fuzzy clustering analysis, it selects the suitable location which is fit for embedding a watermark in color images. Then, the algorithm adaptively changes embedding strength by making use of the visual masking in the wavelet domain, respectively in R, G, B three channels. In the extraction, it does not need the original image. Experimental results show that color images with watermark do not appear any visual distortions, and the algorithm has better robust to some image operations.