Feasibility of underwater true color three-dimensional imaging using hyperspectral LiDAR

Abstract

Hyperspectral LiDAR (HSL) has the advantage of integrating spatial and spectral detection and has become a new direction for underwater three-dimensional (3D) imaging. However, underwater HSL may encounter color distortion caused by the wavelength-selective attenuation property of the water medium. In order to address this problem, we propose the underwater target spectral radiometric correction (UWSRC) method. This method leverages the advantage of LiDAR in the acquisition of the ranging information for each point of the target. Based on the assumption of consistent water transmittance along the laser transmission path, the spectral transmittance of water at each point can be accurately calculated according to the ranging data. Then each point’s spectral reflectance can be corrected, achieving true color reconstruction of 3D point clouds of underwater targets. In order to assess the feasibility of the proposed method, we built a underwater virtual active HSL system to conduct underwater 3D imaging experiments in turbid water environment. The results indicate that the fidelity of the reconstructed colors is significantly improved, and the ranging accuracy has reached millimeter-level. This method can be easily applied to underwater true-color 3D imaging based on HSL and is also expected to be applied to true-color 3D imaging restoration in other transmission medium.