Ground Level Determination in Forested Environment With Utilization of a Scanner-Centered Terrestrial Laser Scanning Configuration

Terrestrial laser scanning (TLS) configurations differ in several ways from airborne laser scanning (ALS) configurations. TLS point clouds have different distance, spatial, and noise distributions than ALS point clouds. Common ground-modeling methods have been first developed for ALS data and are not optimally suited for ground level determination from TLS data. In this letter, we present a ground point labeling method for TLS data. The method uses a priori information of the TLS configurations and scanner-centered coordinates to determine the ground level. The method is tested in a forested scanning scene with six individual TLS scans. The results are compared against ALS reference data. Different parameterizations showed up to 0.25-m bias-corrected root-mean-square deviation from the reference in 15-m range from the scanner, which is within expected variation.