Detecting Inspection Objects of Power Line from Cable Inspection Robot LiDAR Data

Area of Science:

• Electrical Engineering
• Robotics
• Geospatial Analysis

Background:

• Power line infrastructure is expanding into challenging terrains, increasing inspection complexity.
• Traditional power line inspection methods are labor-intensive and difficult.
• Advanced LiDAR technology offers potential solutions for efficient inspection.

Purpose of the Study:

• To develop a novel methodology for detecting inspection objects surrounding power lines using CIR LiDAR and POS data.
• To improve the efficiency and accuracy of power line inspection in complex environments.
• To enable automatic detection of security risks and precise dimensioning of fittings.

Main Methods:

• Data processing involves dividing point clouds into single-span units.
• An optimal elevation threshold is used to remove ground points, enhancing efficiency and accuracy.
• A Structured Partition based on POS data (SPPD) algorithm extracts power lines and surrounding data.
• Partition recognition utilizes local neighborhood statistics and 3D region growing for object identification.

Main Results:

• The proposed method achieves an average accuracy of 90.6% at the point cloud level.
• An average precision of 98.2% is achieved at the point cloud level.
• Experimental results demonstrate the feasibility and promise of the methodology.

Conclusions:

• The developed method effectively detects inspection objects around power lines in complex environments.
• This approach significantly improves the intelligence level of power line inspection.
• The findings support precise modeling and automatic security risk detection.