基于图像分区的置乱算法

Algorithm of Image Scrambling Based on Segmentation

传统的图像置乱加密算法是将图像中的像素位置重新排列,从而使图像变成一幅杂乱无章的新图像。由于像素值并未改变,因此加密图像的灰度直方图并没有改变,而灰度直方图体现了一幅图像的统计特性,因此对加密要求高的领域,这类置乱算法则难以达到期望的要求。针对上述问题,文中提出一种基于图像分区的置乱算法。该算法首先对原始图像进行分区,不同的分区图像采用不同的置乱算法,再对图像的分区进行置乱,并对相邻像素值进行异或运算置乱图像的频率域,选取的置乱算法和迭代次数由Logistic方程产生。实验结果表明,该算法与传统的图像置乱方法相比,置乱速度更快,加密效果更好,安全性更高。

Traditional image scrambling algorithms are accomplished by arranging the pixels position in original image repeatedly, finally a new disorder image is obtained. But the value of original image pixels are not changed, so the histogram after encryption stays still, while the histogram reflects the statistical characteristics of an image,because above reasons only this kind of scrambling algorithm cannot meet the desirable demand in some professional fields with high requirements. Considering all above problems, present a new image scrambling algorithm based on segmentation. First of all, segmenting the original image, the different image scrambling algorithms are ap- plied, then image segments are scrambled, after that, the XOR operation to two adjacent pixels is applied to scramble the frequency do- main, the number of iterations and the selection of scrambling algorithms are determined by the Logistic equation. The experimental results show that compared with some traditional algorithms, this algorithm scrambles faster, has better encryption effect and higher security.