Analysis and Radiometric Calibration for Backscatter Intensity of Hyperspectral LiDAR Caused by Incident Angle Effect.

Wenxin Tian,Yuwei Chen,Miaomiao Pan,Juha Hyyppä,Ziyang Li,Haohao Wu,Lingli Tang,Jing Lu,Peilun Hu,Wenjing He,Jiajia Zhu,Changhui Jiang

doi:10.3390/s21092960

Abstract

Hyperspectral LiDAR (HSL) is a new remote sensing detection method with high spatial and spectral information detection ability. In the process of laser scanning, the laser echo intensity is affected by many factors. Therefore, it is necessary to calibrate the backscatter intensity data of HSL. Laser incidence angle is one of the important factors that affect the backscatter intensity of the target. This paper studied the radiometric calibration method of incidence angle effect for HSL. The reflectance of natural surfaces can be simulated as a combination of specular reflection and diffuse reflection. The linear combination of the Lambertian model and Beckmann model provides a comprehensive theory that can be applied to various surface conditions, from glossy to rough surfaces. Therefore, an adaptive threshold radiometric calibration method (Lambertian–Beckmann model) is proposed to solve the problem caused by the incident angle effect. The relationship between backscatter intensity and incident angle of HSL is studied by combining theory with experiments, and the model successfully quantifies the difference between diffuse and specular reflectance coefficients. Compared with the Lambertian model, the proposed model has higher calibration accuracy, and the average improvement rate to the samples in this study was 22.67%. Compared with the results before calibration with the incidence angle of less than 70°, the average improvement rate of the Lambertian–Beckmann model was 62.26%. Moreover, we also found that the green leaves have an obvious specular reflection effect near 650–720 nm, which might be related to the inner microstructure of chlorophyll. The Lambertian–Beckmann model was more helpful to the calibration of leaves in the visible wavelength range. This is a meaningful and a breakthrough exploration for HSL.

Highlights

Hyperspectral LiDAR (HSL) is a new remote sensing method
To verify tdhieffesrpeencttrwalavanelaelnygsitshsabmileitaysuorfeHd SbLy,twheeHcoSmL pfoarrethdethsaemsppelecst.raTlhreefLleacmtabnecretian–Beckmann measured by HmSoLdewl ipthrotphoastedmienasthuirsepdabpyertihseusSeVdCtospcoercrteroctmtheteebraicnksthcaettceorrirnetsepnosnitdyionfgHSL caused by wavelength ranthgee iinncSidecetniocena3n.1g.leSeecftfieocnt i3n.2Semcatiionnly3a.3n.alyzes the incidence angle effect at different wavelengths measured by the HSL for the samples
We analyzed the relationship between backscatter intensity and incident angle of HSL and introduced the Lambertian–Beckmann model

Summary

Introduction

Hyperspectral LiDAR (HSL) is a new remote sensing method. It has high spatial and spectral information detection ability, which greatly compensates for the disadvantage of the traditional single-wavelength LiDAR in target attribute identification by offering a “colorful” point cloud. In addition to the geometric information, it can record the backscattered optical power of each scanned target at multiple wavelengths. The backscatter intensity of laser at different wavelengths has an accurate one-to-one correspondence with the three-dimensional coordinate information. It has the advantage of pixel-level fusion without registration, which avoids the additional complex data processing process caused by the fusion and registration of traditional single-wavelength LiDAR and hyperspectral remote sensing imaging data. Fine classification of tree species [4] and estimation of plant physiological information (leaf area density [5], leaf water content [6], nitrogen content [7], and chlorophyll content [8,9])

Methods

Results

Discussion

Conclusion