基于分数傅立叶变换和相位恢复算法的彩色图像加密

Color image encryption based on fractional Fourier transform and phase retrieval algorithm

提出一种基于分数傅立叶变换和相位恢复算法的彩色图像加密技术。加密时把待加密彩色图像分解为红(R)、绿(G)、蓝(B)三个分量,并构成联合图,然后将联合图迭代编码到两块相位板中,在迭代过程中,同时改变两块相位板。解密时,只需要将编码得到的两块相位板匹配起来放在正确的位置,就可以恢复出高质量的彩色图像。模拟实验表明,该算法收敛速度快,系统安全性能高,并且可以光学实现。

A novel technique of color image encryption and decryption based on fractional Fourier transfom and phase retrieval algorithm is proposed.The technique decomposes the target color image into three RGB components,and forms a joint image,then encodes the joint image into two random phase mask（RPM） iteratively.This method modifies both the distributions of the two RPM simultaneously in the iteration.In the decryption,a retrieved color image with significantly high quality was obtained when the designed RPM are matched and each of them is located in the correct domain.Simulation experiments show that the algorithm converges rapidly and the system has a high security.This technique is possibly realized by means of optics.