基于模糊聚类的DWT域图像半脆弱水印算法

Semi-fragile image watermarking based on DWT domain and fuzzy clustering theory

提出了一种基于小波变换和模糊聚类的新颖的半脆弱水印算法,该算法将噪声可视函数作为模糊聚类的一个特征,根据聚类结果确定半脆弱水印的嵌入位置和取值。仿真实验结果表明,该算法不仅具有较好的透明性和水印嵌入位置自适应性,对常规图像处理操作具有较好的鲁棒性,而且能对拼贴攻击、替换部分图像等恶意图像篡改进行检测和定位。

A DWT-based novel image semi-fragile watermarking scheme utilizing fuzzy clustering theory is proposed.According to the classification of image’s noise visibility function,gradient etc.we can decide the location and value of watermark.Experimental results show that our semi-fragile watermarking scheme is invisible and is robust to common image processing.The proposed watermarking algorithm is invisible and fragile to content modification and can detect effectively the tampered regions.