Roadside Asset Extraction from Mobile LiDAR Point Cloud

The extraction of traffic signs from Mobile Light Detection and Ranging (LiDAR) point cloud data has become a pivotal focus in research, driven by the growing integration of LiDAR technologies in transportation applications. LiDAR, a remote sensing technology, captures highly detailed three-dimensional point cloud data, offering a comprehensive representation of the surrounding environment. The advantage of efficient data collection for large-scale road networks is particularly evident with Mobile LiDAR systems mounted on vehicles.

This study is centered on developing and refining techniques for extracting traffic signs from Mobile LiDAR point cloud data. The precise detection and localization of traffic signs are crucial for enhancing road safety, navigation systems, and intelligent transportation solutions. The incorporation of LiDAR technology in this context opens up new possibilities for automating the recognition and mapping of traffic signs.

The research specifically delves into the detection of traffic signs using Mobile LiDAR point cloud data. The intensity-based sign extraction method efficiently identifies traffic signs, traffic signals, and other retro-reflective objects, providing valuable insights for transportation asset management. The workflow begins with the management of the LAS dataset, encompassing tasks such as merging/splitting, gridding, and the detection of high-intensity features. Subsequently, the identified signs are placed in Google Earth Pro, enabling their seamless display in Geographic Information Systems (GIS).

Furthermore, the study explores point density analysis, establishing a connection with potential grid resolutions for additional extraction or analysis, such as road condition assessments or crack detection.

Additionally, the research delves into the investigation of Deep Learning point classification and Hough transformation plane detection. The outcomes and limitations of these approaches are comprehensively summarized.