Enhancing Roadway Safety: LiDAR-based Tree Clearance Analysis

TL;DR

This work tackles the problem of ensuring vertical clearance over roadways by automating the detection of vegetation encroachment using street-level LiDAR. It presents an end-to-end pipeline that integrates RandLANet-based semantic segmentation, multi-frame point-cloud concatenation, a novel 2D-contour road boundary algorithm, construction of a clearance-volume up to a regulatory height, and projection of violative points onto images for visualization. Key contributions include a new contour-detection method for the road plane, scalable data fusion across frames, and an actionable clearance gauge with visualization, all accompanied by open-source code and validation on the PandaSet dataset achieving about 93.86% segmentation accuracy. The approach enables municipalities to systematically identify trimming targets, potentially reducing labor and enabling city-wide safety improvements, while acknowledging computational challenges that arise when scaling to large urban areas.

Abstract

In the efforts for safer roads, ensuring adequate vertical clearance above roadways is of great importance. Frequently, trees or other vegetation is growing above the roads, blocking the sight of traffic signs and lights and posing danger to traffic participants. Accurately estimating this space from simple images proves challenging due to a lack of depth information. This is where LiDAR technology comes into play, a laser scanning sensor that reveals a three-dimensional perspective. Thus far, LiDAR point clouds at the street level have mainly been used for applications in the field of autonomous driving. These scans, however, also open up possibilities in urban management. In this paper, we present a new point cloud algorithm that can automatically detect those parts of the trees that grow over the street and need to be trimmed. Our system uses semantic segmentation to filter relevant points and downstream processing steps to create the required volume to be kept clear above the road. Challenges include obscured stretches of road, the noisy unstructured nature of LiDAR point clouds, and the assessment of the road shape. The identified points of non-compliant trees can be projected from the point cloud onto images, providing municipalities with a visual aid for dealing with such occurrences. By automating this process, municipalities can address potential road space constraints, enhancing safety for all. They may also save valuable time by carrying out the inspections more systematically. Our open-source code gives communities inspiration on how to automate the process themselves.

Figures (11)

• Figure 1: Single (a) and concatenated (b) point clouds of sequence 003.
• Figure 2: Point cloud containing only points belonging to the classes road (grey) and vegetation (green) in sequence 011.
• Figure 3: The green point in a) is considered a contour point because one angle exceeds 135° (shown in red). The green point in b) is not a contour point.
• Figure 4: All the road points (a), the sampled road points (b), and the detected contour points in blue.
• Figure 5: The detected points are shown in red (sequence 019).