Research On Image Encryption Algorithm Based On CGI

With the rapid development and progress of computer technology and Internet application,information is confronted with the risk of being compromised.Image information faces serious problems such as maliciously spread,illegal stealing,and illeggal monitoring.Optical information security technology has obvious advantages including high robustness,high freedom,high parallelism as well as high speed,and has been extensively studied.In this thesis,the computational ghost imaging is studied further,besides solving the existing problem,two optical image encryption algorithms are proposed.The details are as follows:1.A sparse reconstruction algorithm of computational ghost imaging based on gradient projection is proposed.First of all,a set of random phase masks is loaded into the space light modulator to illuminate the object,and the transmitted light intensity is recorded by a bucket detector.The random phase masks are compressed and transmitted to the receiver through the secure channel.The receiver performs the sparse reconstruction of these phase masks based on gradient projection,and the original object can be reconstructed by calculating the correlation between the recovered phase masks and the corresponding light intensities.The simulated experiments show that the proposed algorithm has good compression performance and can reduce the number of transmitted masks significantly.When it is not necessary to obtain a high,quality reconstructed image,the existence of the original information can be verified with only transmitting a small number of random phase masks.2.An image encryption scheme based on iterative phase retrieval and computational ghost imaging is proposed.First,a Hadamard matrix with a certail order is generated and each row element is reshaped into a two-dimensional pattern.Then,each pattern is encoded into a phase-only mask using the iterative phase retrieval algorithm.These phase-only masks are loaded into the spatial light modulator sequentially and a bucket detector is used to collect light intensities as the ciphertext.The original image is reconstructed by the correlation calculation between the ciphertext and the corresponding phase-only masks.It can be demonstrated that the reconstructed image with high quality can be obtained only with fewer phase masks.3.An image encryption scheme based on fractional Fourier transform and phase retrieval is proposed.First,the original image is transformed into a complex matrix using the fractional Fourier transform,and this matrix is divided into two parts,i.e.real part and imaginary part,which are the input of the computational ghost imaging system.Then,the light intensities as the ciphertext are collected with a bucket detector.In order to further improve the quality of the reconstructed image,the phase-only masks are generated with the Hadamard matrix formed by using a cake-cutting method.The effectiveness of the proposed scheme is demonstrated with the simulated experiments.