True color 3D imaging optimization with missing spectral bands based on hyperspectral LiDAR

True color 3D imaging plays an essential role in expressing target characteristics and 3D scene reconstruction. It can express the colors, and spatial position of targets and is beneficial for classification and identification to investigate the target material. As a special case of target imaging, true color 3D imaging is important in understanding and reconstructing real scenes. The fusion of 3D point clouds with RGB images can achieve object reconstructions, yet varying illumination conditions and registration problems still exist. As a new active imaging technique, hyperspectral LiDAR (HSL) system, can avoid these problems through hardware configuration, and provide technical support for reconstructing 3D scenes. The spectral range of the HSL system is 431-751nm. However, spectral information obtained with HSL measurements may be influenced by various factors, that further impinge on the true color 3D imaging. This study aims to propose a new color reconstruction method to improve color reconstruction challenges with missing spectral bands. Two indoor experiments and five color reconstruction schemes were utilized to evaluate the feasibility and repeatability of the method. Compared with the traditional method of color reconstruction, color reconstruction effect and color similarity were considerably improved. The similarity of color components was improved from 0.324 to 0.762. Imaging results demonstrated the reliability of improving color reconstruction effect with missing spectral bands through the new method, thereby expanded the application scopes of HSL measurements. (C) 2021 Optical Society of America under the terms of the OSA Open Access Publishing Agreement