Research On Image Encryption Based On Chaotic Mapping And Compressed Sensin

With the rapid development of network and communication technology,information transmission is faster and more convenient.As a new information transmission medium,digital images are widely used in industry,military,medical,national defense and security because of their characteristics of easy generation,storage and image transmission.and other fields.However,these digital images,which contain personal privacy,corporate data and even state secret information,may be deliberately attacked,illegally stolen,modified and destroyed by criminals during network transmission,which will seriously affect personal privacy and corporate interests.and even pose a major threat to national security.Based on this,ensuring the security of digital images has become a research focus in the field of information security.On the basis of existing hardware conditions,while encrypting images,it is also necessary to take into account the compression of redundant data.Regarded as a variant of symmetric encryption,compressed sensing can not only realize simultaneous compression and encryption of data,but also has certain security and good robustness.However,the security of encrypted information using the traditional compressed sensing framework is not strong,and the initial value sensitivity,ergodicity and pseudo-randomness of chaos theory are consistent with the requirements of cryptography.Therefore,this paper aims to combine chaotic systems,compression Perceptual technology designs a safe,efficient,robust,and compressible digital image encryption algorithm.The main work of this paper is as follow.Design a safe and efficient grayscale image compression encryption scheme based on the semi-tensor product compressive sensing model and two-dimensional hyperchaotic mapping.In this scheme,firstly,a controlled cipher stream and measurement matrix are jointly constructed according to the information entropy of the plaintext and the key specified by the user.Then,the hardware-friendly measurement matrix constructed by the linear congruential generator is used to perform synchronous compression sampling on the plaintext sparse coefficient matrix after sparse decomposition,hard thresholding and two-dimensional FAN scrambling.Afterwards,the sampled real number domain coefficient matrix is transformed into the integer domain by means of linear mapping.Finally,under the control of the cipher stream,the compressed image is synchronously scrambled and diffused to conceal the unique statistical properties of the plaintext.In addition,through a series of simulation experiments and security analysis,it can be known that the proposed grayscale image encryption scheme can effectively resist a variety of security attack models,and has good reconstruction performance and a certain degree of robustness.A new fractional order Chebyshev chaotic map is proposed,and its performance is evaluated by bifurcation diagram and Lyapunov exponent.Experimental results show that compared with other newly proposed one-dimensional chaotic maps,the FCCM map designed in this paper has a larger chaotic range and more complex chaotic performance.In addition,on this basis,this paper also proposes a visually secure image encryption algorithm based on the compression-encryption-embedding architecture by combining the two-dimensional compressed sensing model and LSB embedding.In this scheme,under the control of the fractional chaos map,the double encryption of the plaintext image is completed through the measurement matrix,FAN transformation scrambling,chaos diffusion strategy and least significant bit replacement embedding.Finally,through experimental simulation and analysis from the aspects of key,anti-differential attack,imperceptibility and encryption efficiency,it is confirmed that the algorithm proposed in this paper can not only ensure sufficient security,but also reduce the possibility of ciphertext image being attacked.