抗几何攻击的量化鲁棒水印技术研究

提出了一种能够有效抵抗旋转、平移、缩放等基本几何攻击的鲁棒图像水印方案,在其嵌入方案中,提出一种几何不变量——基于图像圆区域内的统计特征不变性。利用该不变性,可以较好地矫正几何形变。在嵌入方案中,实现了在一个嵌入点上嵌入2比特水印信息量,并且利用HVS特性,把水印自适应地嵌入在视觉不敏感区域的纹理方向上。在水印提取方案中,利用特征圆形区域矫正形变,然后在DCT变换域中盲提取水印。实验结果表明,该水印方案具有很好的透明性,同时具有较强的抵抗基本几何形变能力,并且对于一般的图像处理、JPEG压缩等攻击具有较好的鲁棒性。

Research on quantization-based robust watermarking technique against geometrical attacks

This paper proposes a image watermarking scheme efficiently robust to basic geometrical attacks such as rotation，translation and scaling.In the embedding scheme，a geometrical invariable is proposed，which is based on the invariability of statistical feature in the circular image area.Using this invariability，the geometrical distortion is first retrieved.In embedded scheme，the 2 bits watermarking information is embeded in one embedding point.Using HVS，watermarking can be embeded adaptively in the texture direction of visual insensitive area.In extract scheme，feature circular area is used to remedy distortions，and then watermarking is extracted blindly in the DCT domain.Experimental results suggest，the watermarking scheme has the better transparency，at the same time，it has the strong ability in resisting basic geometrical distortions，also it has the better robustness for common image processing，JPEG compression and etc.