Image Compression And Encryption Algorithm Based On Chaotic Systems

With the rapid development of image acquisition technology,the image data and the amount of information it carries are getting larger and larger,and the security in the transmission process and the large bandwidth resource are all problems to be solved.The effective method to solve these problems is to compress and encrypt the image simultaneously.With the characteristics of strong correlation and huge data,the traditional algorithm can't meet the requirements of image compression and encryption.Chaos is a motion that occurs in a certain and nonlinear systems,which seems to be irregular.Chaotic sequences are widely used in image encryption due to their sensitivity to initial values and unpredictability.Based on chaotic system and compressed sensing technology,image compression and encryption algorithm is proposed in this thesis.The main work can be summarized as follows:Firstly,compressed sensing method is used to complete image compression.In order to reduce the amount of data required for transmitting the measurement matrix in the traditional method,a chaotic-based structural randomization measurement matrix is constructed,which took the initial value of chaos as the key,and the feasibility of the measurement matrix was verified by simulation.In the process of image reconstruction,in order to reduce the number of iterations to obtain the optimal value,the adaptive gradient projection reconstruction algorithm is improved based on gradient projection algorithm and nonmonotone line search technology,and the algorithm convergence proof and simulation analysis are carried out to verify the effectiveness of the algorithm.Secondly,the key stream sequence is constructed based on Chen system for image encryption.The chaotic sequences obtained by Chen system were scrambled and rearranged to obtain a three-dimensional key flow with better randomness.Since the data encryption effect obtained by compressed sensing is not so good,the observation value of compressed sensing is used as input,and then multiple rounds of scrambling and diffusion are applied to improve the security of ciphertext.In the process of scrambling,considering the characteristics of data,the one-dimensional scrambling algorithm is improved to make it more suitable for the proposed compression and encryption algorithm.Thirdly,the algorithm is designed.Using the cascaded compressed sensing and scrambling-diffusion framework,the compressed image data after compression sensing is scrambled and diffused to obtain the compressed and encrypted image.The cascading structure makes ciphertext image more robust against various attacks.To verify the effectiveness of the algorithm,numerical simulation and security analysis were carried out from several aspects.Finally,after summarizing the research work of this thesis,The main conclusions of this thesis are given and the direction to be further studied for image compression and encryption technology is pointed out.