Adaptive Image Watermarking Algorithm Based on Fuzzy Cluster in Domain
基于模糊聚类的小波域数字图像水印算法*

Combining the human visual system with the image local properties,an adaptive digital image watermarking scheme based on fuzzy cluster in DWT domain was proposed.First,the DWT was performed on host image.Then,the watermark embedding location and embedding intensity were identified adaptively by applying fuzzy C-means clustering algorithm according to the human visual system in DWT domain.Finally,the watermark was embedded adaptively by modifying the wavelet coefficients according to the image local correlation.The proposed watermarking scheme could extract the watermark without the help from the original image. Experimental results show that the proposed adaptive image watermarking scheme is invisible and robust against various signal processing such as cropping,compression,filtering,noise adding,image enhancement,and mosaic.