| With the rapid development of network and communication technology,more information is transmitted on the Internet platform,and the form of information is gradually diversified.Among them,images are widely used in the Internet of Things,medical care,aerospace.Most of the images transmitted on the network contain sensitive information such as personal privacy or state secrets.Once they are deliberately attacked by criminals,they will pose a great threat to personal information security,company interests and even national security.Therefore,the issue of security and privacy protection in online images has become a research focus in the field of information security.Semi-Tensor Product Compressed Sensing(STP-CS)theory can compress and encrypt digital image simultaneously.It breaks the dimension matching condition that traditional matrix multiplication needs to meet,and reduces the measurement matrix space of encryption algorithm.Moreover,the corresponding parallel reconstruction algorithm can greatly improve the decryption speed and is an effective method to protect the image content.Therefore,it is of great research value to consider adopting STP-CS to encrypt digital images in the network.However,the existing literature on STP-CS method for image encryption still has the problem of low security.Aiming at solving these problems,this paper first designs a novel Coupled Map Lattice(CML)system.On this basis,an efficient image encryption algorithm is proposed.The algorithm adopts the structure of “scrambling,compression and diffusion”.First,we adopt Arnold scrambling to disrupt and equalize the distribution of sparse coefficients.Then,we apply STP-CS for compression and preliminary encryption,and finally DNA diffusion is designed to strengthen the encryption effect.Simulation results and security analysis show that the encryption system has a large key space.It is highly sensitive to keys,and has strong resistance to chosen plaintext attacks,statistical attacks,and differential attacks.Besides,it is resistant to noise and data loss attacks with good robustness.All these excellent encryption performance promote the safe transmission of images on public channels and networks.In addition,this article also extends the encrypted object to multi-image scenes.Based on bit-plane recombination and DNA-plane recombination technology,a doubleimage compression encryption algorithm based on STP-CS is proposed for the first time.In this algorithm,first,two images are adaptively thinned to improve the quality of decryption.Subsequently,the chaotic index sorting with more stable sorting performance is used to scramble the sparse coefficients of the two images respectively.Next,STP-CS is used to simultaneously compress and initially encrypt the two image coefficients.Finally,for the two quantized intermediate cipher images,bit plane recombination and DNA plane recombination are used to disrupt the plaintext related information left by the intermediate cipher images to ensure the security of encryption.Experimental results show that the encryption scheme has a larger key space,higher plaintext and key sensitivity,stronger anti-statistical analysis ability,and better robustness.During decrypting phase,the quality of the decrypted image and the speed has also been effectively improved.This is of great significance to the privacy protection of images in network transmission. |
