Fund Project:Project supported by the National Natural Science Foundation of China (Grant Nos. 61875093, 61465005) and the Natural Science Foundation of Hebei Province, China (Grant No. F2018402285).
Received Date:25 December 2018
Accepted Date:19 March 2019
Available Online:01 June 2019
Published Online:05 June 2019
Abstract:In this paper, a new method of encrypting a color image based on θ modulation is proposed by using the tricolor principle and computer-generated hologram (CGH) technology. The encryption process includes the θ-modulated three primary color components and the coding of computer-generated hologram, which is implemented in a Fresnel diffraction and spatial filtering system. Firstly, the color image modulated by the first random phase key is divided into three encryption channels by red laser, green laser, blue laser, and tricolor filters. Each channel is introduced by a transmissive amplitude-type sinusoidal grating with different directions, which is used to separate the three primary color components in the spatial spectrum plane. Secondly, the modulation results of tricolor components are superimposed together to form a compound image, and the phase truncation of the superposition result is performed to achieve the asymmetric encryption. Finally, the amplitude of the compound image is modulated by the second random phase key and is encoded into a binary real-value gray-color CGH by Roman-type coding method. Therefore, the gray-color information of the original image is completely hidden in the encrypted CGH, which is more general and deceptive in the storage and transmission process. Decryption is an inverse process of the encryption. Firstly, the encrypted CGH is placed on the input plane of the spatial filtering and Fresnel diffraction system. Secondly, the demodulation of CGH phase key and the spatial filtering based on optical filter are performed. Finally, the color plaintext image is obtained by using the correct Fresnel diffraction. The simulation results show the validity and feasibility of the proposed method. In addition, the anti-noise attack and anti-shearing attack performance of this color image encryption method are investigated. Compared with results from the three presented methods reported in the literature, our investigated results demonstrate that this method has good robustness to noise attack and shearing attack, and has obvious advantages when the attack noise density is larger. Due to the characteristics of high security and anti-noise, we believe that this color image encryption method promises to have important applications in the information transmission and multi-user authentication. Keywords:color image encryption/ computer generated holography/ θ modulation/ asymmetric encryption
全文HTML
--> --> -->
2.加密过程本文提出的彩色图像的加密过程分为三基色分量θ调制加密和计算全息编码两个步骤, 第一个步骤在图1所示的菲涅耳衍射和θ调制的4f系统中完成. 图 1 彩色图像的菲涅尔衍射和θ调制系统 Figure1. Fresnel diffraction and θ modulation system for color image.
解密时, 设置滤波器分别只恢复彩色图像的R, G和B三个分量, 如图8(a)—(c)所示, 图8(a)—(c)是相应的原始三基色分量, 它们对应的相关系数分别为${\rm{CC}} = 0.876$, ${\rm{CC}} = 0.891$, ${\rm{CC}} = 0.903$, 图像逼真度分别为${\rm{IF}} = 0.881$, ${\rm{IF}} = 0.900$, ${\rm{IF}} = 0.911$. 从图8可以看出, 恢复出来的图像丢失了部分细节信息. 造成该现象的原因有两个: 一是由于第一个随机相位密钥的散射作用使得部分高频分量在截取傅里叶频谱时被舍去; 二是加密结果的计算全息量化编码引入了一定误差. 经验证, 当降低第一个随机相位密钥的动态取值范围并增加计算全息量化编码的像素点数时, 解密图像的质量会出现很大提升. 因此, 原始图像的灰度和色彩信息可以被恢复出来, 其恢复质量在一定程度上取得了令人满意的效果. 图 8 解密彩色图像三基色分量 (a)红色分量; (b)绿色分量; (c)蓝色分量; (a'), (b'), (c')原始图像三基色分量 Figure8. Tricolor components of color image: (a) Red component; (b) green component; (c) blue component; (a'), (b'), (c') red, green and blue components of original color image.
24.2.抗剪切攻击测试 -->
4.2.抗剪切攻击测试
本文还针对加密图像做了抗剪切和抗噪声模拟实验, 图9给出了对加密图像做不同程度剪切后的实验结果和经剪切攻击后恢复得到的解密图像. 图 9 不同程度的剪切攻击和剪切攻击后的恢复图像 Figure9. Shear attack of different levels and image restoration after shear attack.