Research On Key Technologies Of Hyperspectral Image Multi-Parameter Compression And Encryption

The era are altering.Technology continues to advance.Optical image encryption technology has become a widely focused focus of scholars.For hyperspectral images,it occupies an important position in the field of military medical and commercial value remote sensing,but its optical encryption technology has not been deeply studied.In addition,it has large amount of redundant data,narrow and short bands,high dimensionality.Therefore,the information value is more.At the same time,it has spatial structure,high correlation between spectra and high degree of connection between bands are also high.It will cause the increase of data redundancy and limit the security range of storage and transmission.In response to the above problems,this paper studies the key technologies of hyperspectral image Multi-parameter compression and encryption.Make the system safe and effective without loopholes.On the basis of ensuring the security level,the amount of data is reduced,so that the hyperspectral image can be obtained safely by the recipient.The relevant research contents are presented as follows:Firstly,a algorithm based on DWT is introduced.The discrete Walsh transform is introduced to compress and reconstruct the hyperspectral image for analysis,and compare it with some compression image processing algorithms at this stage.This algorithm divides the hyperspectral input image into single-band processing and compresses it through methods such as block processing and zeroing of low-frequency coefficients,which can easily achieve the purpose of quantitative control of the compression ratio,and the realization is relatively easy,which facilitates the storage and transmission of data information.Secondly,a DWT-based multi-parameter compression fingerprint encryption algorithm for hyperspectral images is proposed.First,perform compression processing on the original hyperspectral input image,then perform multi-parameter Fourier encryption transformation on the compressed image,and combine it with fingerprint recognition encryption technology,one of the biometric recognition technologies,to perform a reinforced two-level encryption system processing.When the keys are correct,if the matching is successful,transform the obtained image such as phase truncation.This scheme has realized the expansion of the relevant key space,the rapid data transmission and storage under the condition of quantitative control of cr,and the security performance has been further improved.Finally,a multi-parameter Fourier domain PEMD encryption algorithm for compressed hyperspectral images is proposed.By performing PEMD processing on the original hyperspectral input image,the asymmetric ciphertext obtained by the asymmetric encryption method is obtained.The use of this structure achieves the purpose of asymmetric encryption output,and can correspondingly perform real-value output.The quality of image reconstruction is maintained at a high level,while the amount of encrypted data is reduced.The access efficiency and transmission rate have been increased to a certain extent,and its ability to resist statistical attacks has been correspondingly enhanced.Its ability to resist statistical attacks has also been enhanced accordingly.