Research On Hyperspectral Image Degredation And Image Quality Enhancement In Near Space Platform

The near space detection system is capable of long-term fixed-point hovering.Carrying optical detection load on it can realize regional continuous observation and wide range stereo monitoring for thousands of kilometers.It has become an important means to achieve remote detection,rapid detection,tracking and identification of small and dim targets in the air.Hyperspectral load can provide more detailed spectral information of target,and has become a hot spot in the detection and recognition of dim targets in complex environment.However,under the condition of observation in near-space,the interference of atmospheric absorption and path radiation on the spectral characteristics of the target is very serious,which will lead to a significant decrease in the detectability of the target.Because of the low spatial resolution of target,the target is mixed with the environmental background radiation.This will lead to the reduction of the accuracy of target detection and recognition.Aiming at this problem,this paper carries out research on the degradation of image quality in near-space,atmospheric correction method and hyperspectral image super-resolution reconstruction method,and carries out corresponding experimental verification.The main research work and results are as follows:(1)Research on degradation mechanism and model of hyperspectral imaging in near-space.From the whole link of hyperspectral imaging in near-space,this paper focuses on the analysis of the absorption and scattering characteristics of the main components of the atmosphere in near-space and the influence at different observation geometric conditions on the path radiation and transmittance of the atmosphere.The influence of link factors such as hyperspectral imaging and adjacent space platform on image quality is also studied.Then,this paper established the model of near-space detection signal transmission,atmospheric radiation transmission,hyperspectral detection system and platform vibration.It provides support for subsequent research on near space atmospheric correction and hyperspectral image reconstruction.(2)Atmospheric correction method for near space based on atmospheric parameter inversion is proposed.On the basis of the study of the traditional atmospheric correction algorithm,the high precision atmospheric transmission simulation model is used to study the inversion of water vapor content based on Atmospheric Precorrected Differential Absorption(APDA)method and visibility based on Dense Dark Vegetation method.The iterative optimization study of the two parties is carried out to ensure the accuracy of water vapor content and visibility.On this basis,the accurate calculation of atmospheric correction parameters is reali zed.The mapping relationship between water vapor content and APDA,visibility and apparent radiance,water vapor content,visibility and atmospheric correction parameters is constructed so as to realize rapid atmospheric correction in near-space.(3)Hyperspectral image reconstruction method based on spectral mixture model is proposed.According to the quality degradation mechanism of hyperspectral image,combined with spectral mixture model,the endmember abundance "sum-to-one" and "non-negative" are the constraints,a super-resolution reconstruction model combining spatial dimension and spectral dimension information of high spatial resolution images and hyperspectral images is constructed under the constraint of "sum-to-one" and "non-negative".Combined with gradient projection method,this paper studies the efficient solution algorithm of the model,thus realizing super-resolution reconstruction of hyperspectral images and providing support for the unmixing and identification of effective spectral features of small and weak targets in near-space.