LiDAR-Based Urban Three-Dimensional Rail Area Extraction for Improved Train Collision Warnings

Area of Science:

• Urban rail transit safety
• LiDAR-based obstacle detection

Background:

• Object intrusion into track areas poses a significant safety risk for urban rail transit.
• Current LiDAR-based obstacle detection methods struggle with sparse data, limiting track area segmentation to short distances.

Purpose of the Study:

• To develop a novel approach for accurate and long-range track area extraction from LiDAR data.
• To overcome the limitations of existing methods in segmenting track regions for urban rail safety.

Main Methods:

• Indirectly extracting track areas by detecting parallel reference structures (e.g., walls).
• Utilizing reference point selection and curve fitting to generate track reference curves.
• Extrapolating a centerline, expanding it to a 2D area, and acquiring the 3D area by ground detection and point filtering.

Main Results:

• The proposed technique successfully extracts track regions from LiDAR data over significantly longer distances compared to conventional algorithms.
• Effective performance demonstrated across diverse urban rail environments including tunnels, elevated sections, and level tracks.

Conclusions:

• The developed method provides a robust solution for long-range track area extraction in urban rail transit.
• This advancement is crucial for improving the reliability of obstacle detection and collision avoidance systems.