Accuracy Assessment of Point Cloud from Mobile Mapping Systems applied to Feature Mapping on Road-surface and Road-side

Abstract

This article investigates the evaluation of point clouds from three Mobile Mapping Systems (MMS) by categorizing methods into two groups: 1) assessing general product quality (positional accuracy and point density) via comparison with Ground Control Points (GCP) measured by Real-Time Kinematics (RTK), and 2) applying feature assessments to evaluate the positional accuracy of road features (lane markings, traffic signs, billboards) against the HD Map standard, while considering compliance with relevant standards and regulations. Linear feature positioning accuracy is assessed using similarity comparisons of feature alignments based on point-wise coordinate differences and feature-to-feature distances (Hausdorff, Euclidean, and Fréchet). The study observed a potential negative correlation between vehicle speed and positional accuracy/point cloud density. The MMS-II system shows mapping capabilities comparable to the MMS-I system, but the MMS-III system exhibits abnormal behavior requiring further verification for mapping applications. The main purpose is the application of the TAICS TR-0016(E) V1.0:2021 HD map standard to support smart city engineering.