LiDAR-Based Real-Time Detection and Modeling of Power Lines for Unmanned Aerial Vehicles

Area of Science:

• Electrical Engineering
• Robotics
• Geospatial Analysis

Background:

• Growing global electricity demand necessitates efficient power line monitoring.
• Traditional inspection methods (foot patrols, helicopters) are costly and risky.
• Unmanned Aerial Vehicles (UAVs) offer a safer, more cost-effective alternative for inspections.

Purpose of the Study:

• To develop a reliable and autonomous power line detection and modeling algorithm.
• To overcome limitations of visual inspection methods in adverse conditions.
• To introduce a novel approach using LiDAR for power line analysis.

Main Methods:

• Proposed the Power Line LiDAR-based Detection and Modeling (PL2DM) approach.
• Utilized scan-by-scan adaptive neighbor minimalist comparison on point cloud data.
• Modeled power lines horizontally as straight lines and vertically as catenary curves.
• Grouped and matched line segments based on collinearity properties.

Main Results:

• The PL2DM algorithm demonstrated promising results on a real-world dataset.
• Achieved efficient processing times, enabling real-time capabilities.
• Successfully detected and modeled power lines using LiDAR data.
• Provided real-time object-based perception for further processing.

Conclusions:

• LiDAR-based power line detection and modeling is a viable and effective approach.
• The developed algorithm enhances the reliability and autonomy of power line inspections.
• The method offers significant improvements over traditional visual inspection techniques.
• Enables real-time perception for advanced power infrastructure monitoring.