基于不变矩和Hilbert码的矢量居民地零水印算法

提出了一种针对矢量居民地数据的零水印算法。为增强零水印的抗裁剪能力,算法利用Hilbert排列码对数据分块以便构造多个零水印。每块零水印的构造方法为:首先,利用Arnold技术置乱原始水印图像;然后,计算居民地图形的矢量不变矩集并将其映射成一个二值矩阵;最后,将置乱后的水印图像与不变矩集所映射的二值矩阵进行异或运算,完成零水印构造。实验结果表明,该算法能够抵抗常见的平移、缩放、旋转等几何攻击,对裁剪攻击也具有较好的鲁棒性。

A Zero-Watermarking Algorithm Based on Invariant Moments and Hilbert Codes for Buildings

A zero-watermarking algorithm for vector buildings data was put forward. Multiple zero-watermarks were constructed by using the Hilbert permutation code to enhance the performance of resisting cropping. The construc-tion method for each block of zero-watermark was to be as follows. Firstly, the Arnold method was employed to shuffle the original watermark image. Then the vector invariant moments of buildings were calculated and mapped to a binary matrix. Finally, an XOR operation was done between the shuffled watermark image and the binary ma-trix mapped by invariant moments of buildings, and the zero-watermark image was obtained. The experiments show that the algorithm can resist common geometric attacks including translation, scaling and rotation and it also has good robustness aggainst cropping attack.