Research On Multi-parameter Encryption Technology Of Hyperspectral Image

In recent years,optical image encryption technology has received extensive attention from scholars due to its multiple degrees of freedom such as wavelength,amplitude,and phase.However,most of the existing optical encryption systems are designed for grayscale,color,or multiple images.As the important element of military and commercial remote sensing,the optical encryption technology of hyperspectral images has not been studied in depth.In addition,the characteristics of hyperspectral images with multiple bands and large data amount have restricted them storage and transmission.In order to solve the above problems,this paper studies the multi-parameter encryption technology of hyperspectral images,which improves the security of the encryption system and reduces the data amount of key and encrypted image,thereby realizing the fast and safe transmission of hyperspectral images.The main research contents of the paper include:Firstly,the hyperspectral image multi-parameter encryption algorithm based on hybrid chaotic mapping is proposed.The hybrid chaotic mapping is introduced to scramble the hyperspectral image,and then the chaotic random phase mask is generated to modulate the scrambled image.Finally,the phase-truncated discrete multi-parameter fractional Fourier transform is performed to obtain the final encrypted image.This algorithm introduces the phasetruncated discrete multi-parameter fractional Fourier transform into the hyperspectral image encryption,which can realize asymmetric encryption while expanding the key space,and the security is significantly improved.In addition,the scrambling operation and generation of the random phase plate are implemented by using the hybrid chaotic mapping with better chaotic performance.They make the additional keys only need to transmit multiple parameters to realize multi-parameter encryption,which is convenient for storage and transmission.Secondly,the hyperspectral image multi-parameter encryption algorithm based on phasetruncated binary tree is proposed.The hyperspectral image is first chaotically scrambled,then each band of the scrambled images is input into the phase-truncated binary tree structure and the encryption process is implemented along the channels from the leaf nodes to the topmost branch node.The binary tree structure is based on chaotic phase mask and multi-parameter fractional Fourier transform,which realizes multi-parameter encryption and expands the key space at the same time.Moreover,it enables different band images have different keys,encryption and decryption paths further improving security.In addition,the tree structure can make the final encrypted data have only 25% of the original image while ensuring high decryption accuracy,saving storage and transmission space.Finally,the multi-parameter encryption algorithm for hyperspectral images based on equal modulus decomposition tree is proposed.The hyperspectral image is chaotically scrambled and input into the equal modulus decomposition tree structure for processing,and then asymmetric ciphertext can be obtained.Asymmetrical encryption and real-value output can be realized by using the equal modulus decomposition structure,and an equal modulus decomposition tree is constructed to implement multi-parameter encryption,which can reduce the amount of encrypted data greatly while maintaining high image reconstruction quality,improving data storage and transmission effectiveness.Moreover,different encryption parameters and paths are provided for different band images,which improves algorithm security.In addition,this algorithm has enhanced resistance to statistical attacks compared with the above algorithm based on phase-truncated binary tree.In this paper,these above algorithms are simulated and analyzed by computer.All of them can realize multi-parameter encryption and real-value output.They have a high level of security and are convenient for storage and transmission,which have good application prospects.