Using Lidar data to enrich the diagnosis of safety problems and collision causes

The current practices of diagnosing safety problems are associated with many challenges such as the reliance on manual observations, intra- and inter-observer variability, time consumption, and the great effort required to conduct a large-scale diagnosis of an entire road network. This research advocates using LiDAR data to create an accurate 3D model of crash-prone locations which would help identify potential safety problems in a robust and efficient manner. To diagnosis safety issues, different algorithms are used to extract and evaluate roadway features such as available sight distance, horizontal and vertical curves characteristics, cross-section elements, and lateral placement of roadway signs.