基于三维Hilbert曲线扫描的图像置乱

Image Scrambling Based on Three-Dimensional Hilbert Curve Scanning

Hilbert曲线的空间填充性质使得它可用于图像置乱.针对二维Hilbert曲线扫描波动性相对较小以及少数几次置乱后仍达不到较好置乱效果的弱点,提出将二维图像堆积成三维矩阵,并利用三维Hilbert曲线产生的扫描矩阵进行图像置乱.数值实验结果表明,三维Hilbert曲线扫描的置乱效果比二维好,算法对于剪切、JPEG压缩、噪声、涂鸦等攻击具有较好的鲁棒性.

The space-filling characteristic of Hilbert curve can be utilized to scramble image. In order to overcome the weakness of two-dimensional Hilbert curve, such as the relatively small undulation and bad scrambling effect, an image scrambling scheme is proposed in which two-dimensional image is piled to be a three-dimensional matrix and the three-dimensional Hilbert curve is applied to generate one scanning matrix to scramble the image. Experimental results show that the scrambling effect based on three- dimensional Hilbert curve is better than that based on two-dimensional Hilbert curve. The robustness of the proposed scheme is tested as well.