Experimental Study of Spectral Radiometric Correction for Underwater 3D-Imaging by Hyperspectral LiDAR
Year: 2026
Type: article
Abstract: Hyperspectral LiDAR (HSL) has emerged as a cutting-edge technique for underwater 3D imaging. Nevertheless, wavelength-selective absorption and scattering of light by water induce severe distortion in the spectral information of submerged targets and further result in color casts. In this paper, the underwater target spectral radiometric correction (UWSRC) method is adopted to correct the spectral reflectance of underwater targets by combining pre-measured water spectral transmittance and range data acquired by LiDAR. This approach obviates the complicated in situ measurement of water inherent optical parameters, which reduces the operational complexity for field deployments of underwater hyperspectral LiDAR remote sensing. Its spectral correction capability is experimentally validated under varying detection distances and diverse aquatic environments, including clear tap water and eutrophic turbid water. Furthermore, a dedicated HSL prototype for underwater scanning and a semi-submerged pod are developed. With this prototype, the 3D scanning imaging distance can be extended to 5.3 m in an outdoor swimming pool. Experimental results demonstrate that the UWSRC method can effectively compensate for the spectral reflectance of underwater targets under different water qualities and detection ranges. Meanwhile, true-color 3D point clouds of a color-calibration board are successfully obtained during long-distance scanning. Error analysis of reflectance inversion is carried out to explore the sensitivity to water transmittance bias and ranging uncertainty. The experimental investigations in this paper verify that the UWSRC method achieves favorable performance in underwater target spectral restoration and true-color reconstruction across different water qualities and detection distances.
Source: Remote Sensing
DOI: 10.3390/rs18223927
Authors: Yicheng Wang(Conceptualization, Methodology, Software, Writing – original draft, Writing – review & editing), Huifeng Tao(Data curation, Formal analysis), Tingting He(Formal analysis, Visualization), Yunhui Guo(Methodology, Supervision), Hao Zhang(Resources, Visualization), Boyu Liu(Formal analysis), Shouzheng Zhu(Resources), Haifeng Li(Data curation), Hao Pan(Conceptualization, Methodology, Project administration, Software, Visualization)
Language: English
Subfield: Forestry
Domain: Life Sciences
Keywords: hyperspectral LiDAR, color cast correction
Open Access status: gold
Estimated APC paid: $3,258
OpenAlex ID: W7233901625
