Abstract: Combining the human visual system with the image local properties, this paper proposes a watermarking algorithm based on wavelet transform and Fuzzy C-Mean(FCM) clustering. Watermarking embedding locations are identified adaptively on the wavelet domain coefficient of the B channel by applying FCM clustering algorithm based on the human visual system. And the watermarking is embedded adaptively according to the image local correlation and cluster result. The watermarking can be extracted without the help from the original image. Experimental results show that the watermarking scheme is invisible and robust against various signal processing such as JPEG compression, image enhancement, cropping, noise adding and filtering.

Key words: B channel, wavelet transform, Fuzzy C-Mean(FCM) analysis, blind watermarking

摘要： 利用人类视觉系统(HVS)的掩蔽效应和图像的局部相关特性，提出一种结合小波变换和模糊聚类分析的水印嵌入算法。在图像B通道下的小波域系数中，根据HVS的掩蔽性进行模糊聚类分析以选择适合嵌入水印的区域，结合图像的局部相关性和隶属度自适应地嵌入水印。算法在提取水印时无需原始载体图像的辅助。实验表明，该算法具有较好的透明性，而且对诸如JPEG压缩、图像锐化、几何裁剪、叠加噪声和平滑滤波等攻击均具有较好的鲁棒性。

关键词: B通道, 小波变换, 模糊聚类分析, 盲水印

CLC Number: 

• TP391