Analysis and exploration of spatial and spectral information with small-footprint hyperspectral lidar returns

Full-waveform hyperspectral light detection and ranging (FWHSL) is vital in retrieving spatial and spectral information during laser scanning. Four main influence factors, the distance between two neighbor targets, the coverage ratio, the incident angle, and the target's reflectance, determine the information of the FWHSL returns. Previous studies mainly focus on the influence of the neighbor distance, incident angle, and reflectance, while we focus on the coverage ratio of the targets in a laser footprint. We propose a novel multispectral waveform decomposition method, including the Trust-Region algorithm for single wavelength waveform decomposition, 3σ rule for screening decomposition results and correction between multispectral waveform decomposition, to obtain the accurate spatial and spectral information from the multispectral returns, which realizes the decomposition error less than 0.3cm when the neighbor distance is 40cm, for a 4ns pulse width LiDAR signal. We find the intensity and overlapped ratio of the returns are strongly related to the coverage ratio, which may accelerate the progress in point cloud information extraction and target recognition.