航天器工程2024，Vol.33Issue(1)：75-83,9.DOI:10.3969/j.issn.1673-8748.2024.01.011
基于辐射不变点的高光谱遥感传感器交叉定标方法
Cross-calibration Method of Hyperspectral Remote Sensing Sensor Based on Radiation Invariant Points
李映潭 1杨辽 2王杰1
作者信息
- 1. 西华师范大学 地理科学学院,四川南充 637002
- 2. 中国科学院新疆生态与地理研究所,乌鲁木齐 830011
- 折叠
摘要
Abstract
In addressing the radiometric calibration issue in remote sensing image data processing,the key lies in accurately obtaining radiometric calibration coefficients to ensure a precise correla-tion between image digital number values and terrestrial radiative values.The article introduces EMIT L1B data and building upon previous experiences and methods from spectral imaging in-struments,proposes a cross-calibration method for hyperspectral remote sensing sensors.This method,utilizing multi-satellite data and automated processes,acquires calibration coefficients for hyperspectral satellites and enhances the accuracy and efficiency of invariant point selection for ra-diometric calibration.Utilizing the intersection points of data collected by EMIT and the ZY-1-02D satellite,accurate calibration coefficients were derived through regression analysis at ra-diometrically invariant points.Simultaneously,Sentinel-2 data are introduced for selecting radio-metric invariant points,employing an automated program to compare the similarity between two spectra.This approach further enhances the efficiency and reliability of point selection.The cali-bration results are evaluated using standard errors and correlation coefficients to assess errors and correlations between bands.The maximum standard error is found to be below 0.008,with 145 band correlation coefficients exceeding 0.90.This demonstrates that the cross-calibration method can yield high-quality reflectance spectral curves,underscoring its importance in improving the ac-curacy and usability of image data.关键词
交叉定标/资源一号02D卫星/地表矿物粉尘源调查卫星/高光谱Key words
cross-calibration/ZY-1-02 D satellite/EMIT/hyperspectral分类
信息技术与安全科学引用本文复制引用
李映潭,杨辽,王杰..基于辐射不变点的高光谱遥感传感器交叉定标方法[J].航天器工程,2024,33(1):75-83,9.
