Terrestrial hyperspectral light detection and ranging (LiDAR) (HSL), as one novel LiDAR system integrated with more wavelengths, has supplied enhanced possibilities in many fields by delivering the geographic coordinate and the hyperspectral backscattered intensity simultaneously. However, like the traditional single-wavelength LiDAR, it is faced with the distance effect, which restricts the accurate acquisition of the hyperspectral intensities and the quantitative retrieval of characteristics of the targets. The previously reported distance effect correction models mainly concentrated on the distance effect correction for the single-wavelength LiDAR system, and some normalization methods with the standard reflectance panel were also provided to calibrate the data measured by multispectral LiDAR and HSL, producing the reflectance-related value. Nevertheless, it is valuable to systematically explore the distance effect of HSL over long distances and search for the corresponding correction methods. Along these lines, in this study, the homogeneous targets and standard reflectance panels with different reflectance values were employed to participate in the distance experiment to systematically explore and analyze the distance effect phenomenon on terrestrial HSL data. A novel practical piecewise fitting model was proposed to output realistic correction results and it was also compared with the classic inverse distance square law and polynomial fitting model. To further demonstrate the feasibility and good capability, the models were employed in the intensity correction for the various vegetation leaf species. A simple 3-D scene experiment was then carried out to demonstrate the intensity variation before and after the implementation of the distance effect correction. From the acquired result, it suggests that (1) the distance effect on terrestrial HSL data is not related to the wavelength and the scanned target. All the wavelengths share a common distance effect; (2) the concept of the ‘distance effect function’ could represent well the distance effect for the terrestrial HSL; (3) as far as the correction ability is concerned, the proposed piecewise fitting model achieved better performance than the others because it exhibited the smallest standard deviation (σ), coefficient of variation (CV), and the ratio of the coefficient of variation (ε) for the correction results of the homogeneous targets, and the leaf experiments confirmed it again; and (4) the further 3-D experiment result exhibited the intensity variation. It proves that the distance effect does exist in the 3-D scanning results and the correction is necessary. Besides, for small-scanned scenes, the incidence angle effect dominates and largely affects the intensity data instead of the distance effect. This study enriches the study of the distance effect mechanism on the terrestrial HSL data, which facilitates the development of research on the radiative effect correction of the HSL. At the same time, the proposed model can be adopted for the analysis and correction of the distance effect on terrestrial HSL data in actual scanning tasks.
Read full abstract