Research And Realization Of Mid-infrared Hyperspectral Image Simulation

Remote sensing image simulation is based on the analysis and understanding of the imaging process of the remote sensor.The imaging process of the designated remote sensor is simulated through a digital model,and then the required remote sensing image is generated.It can provide an important theoretical basis for the feasibility analysis and demonstration,design and manufacture,operation and maintenance,and later management guarantee of the remote sensor load design process,thereby ensuring the efficiency of the remote sensor design.In recent years,the rise of hyperspectral remote sensing technology has played a huge role in many fields.However,due to the lack of hyperspectral image acquisition and the high launch cost of spaceborne hyperspectral satellites,hyperspectral image simulation research has emerged.In the visible and near-infrared bands,remote sensing image simulation technology has achieved many results.Unlike the visible light channel,the energy detected by the mid-infrared channel sensor includes both the reflected solar radiation energy and the thermal radiation emitted by the ground object itself.Therefore,the mechanism of the mid-infrared channel is complicated,the simulation is difficult,and the research work of image simulation is relatively lacking.In response to the above problems,this paper carries out different methods of mid-infrared channel hyperspectral image simulation research.The specific research content is as follows:(1)Based on the combined image simulation method of virtual reality and remote sensing mechanism,the mid-infrared hyperspectral image is simulated.This method simulates the entire chain of remote sensing imaging to simulate mid-infrared hyperspectral images of the threedimensional scene,which can be imaged according to the requirements.Conditions,artificially and quantitatively control simulated surface parameters,atmospheric conditions and other parameters to generate mid-infrared hyperspectral remote sensing data.(2)Based on the correlation characteristics of the adjacent channel spectra of remote sensing images,analyze the correlation between the data source channel and the channel to be simulated,and build a data conversion model of the surface emissivity of the channel to be simulated on the basis of statistics and analysis,and the accuracy of the model is analyzed,and the average relative error of the model is less than 10%.(3)Use the constructed mid-infrared channel conversion model to simulate the surface emissivity data of the mid-infrared channel.First,process the data of Fengyun 4 to eliminate the invalid value area to achieve the purpose of accurate and credible simulation data.Improve the accuracy and overall effect of simulation data.Realize the generation of continuous hyperspectral surface emissivity data in the mid-infrared band of 2.5-5.0?m and a spectral resolution of 50 nm.(4)Software design and implementation of the mid-infrared image simulation method based on remote sensing images,and build a user-interactive mid-infrared data simulation software.The software is divided into data query module,surface data simulation module,and data display module according to function and purpose.The four main modules of the data management module and the data management module realize the remote sensing image simulation in the mid-infrared band.In summary,this thesis systematically researches mid-infrared hyperspectral image simulation,using imaging simulation methods based on the combination of virtual reality and remote sensing mechanisms,and remote sensing image channel conversion models to perform mid-infrared hyperspectral images.Simulation effectively solves the difficult problem of midinfrared waveband remote sensing image acquisition,provides theoretical support for mid-infrared waveband image simulation and load development applications,and at the same time develops image simulation software,which provides a powerful means for mid-infrared image simulation.The paper has 38 diagrams,16 tables and 81 references.