•  
  •  
 

Journal of System Simulation

Abstract

Abstract: To achieve accurate reconstruction of thermal infrared hyperspectral images under limited spectral bands, this paper proposed a reconstruction method based on physical modeling and simulation. Semi-global decomposition algorithm was adopted to invert the thermophysical properties of the scenario based on the physical model of thermal radiation, simulating and generating full-band hyperspectral data. An optimal spectral band selection strategy driven by a physical model was proposed, which integrated the sensitivity of temperature inversion and the separability of material spectra. Experiments were conducted on both simulated and measured datasets to evaluate the performance of material identification, temperature inversion, and spectral reconstruction under different numbers of spectral bands and selection strategies. Simulation results show that the proposed strategy improves the inversion accuracy of hermophysical properties when the number of spectral bands is small. In terms of spectral reconstruction, the strategy effectively suppresses local error peaks and improves the overall stability and reliability of reconstruction by ensuring the simulation accuracy of critical wavenumber regions.

First Page

3189

Last Page

3201

CLC

TP751.1

Recommended Citation

Yang Yonghao, He Xiaoyu. Analysis of Optimal Spectral Bands for Thermal Infrared Hyperspectral Image Reconstruction Driven by Physical Simulation Model[J]. Journal of System Simulation, 2025, 37(12): 3189-3201.

Corresponding Author

He Xiaoyu

DOI

10.16182/j.issn1004731x.joss.25-FZ0720

Download

Share

COinS