An Improved Algorithm for Foreign Objects Detection on Power Transmission Lines

Da-Ming Zou,Yangkai Liu,Hang Lin,Zhen Qiu

DOI: https://doi.org/10.1109/icsai61474.2023.10423329

2023-01-01

Abstract:Foreign objects threaten transmission line reliability and can cause cascading blackouts. Manual inspection of thousands of miles is inadequate. This paper presents an optimized deep learning model for real-time foreign object detection. We enhance YOLOv4 with GhostNet, CBAM, and BiFPN to to simplify the model structure and improve efficiency and accuracy. GhostNet enables lightweight detection through efficient convolutions. CBAM and BiFPN refine representations. Experiments show state-of-the-art accuracy while using 4.4x fewer parameters than YOLOv4. Our model operates in real-time on edge devices, advancing automated inspection to monitor critical infrastructure. Precise and timely foreign object detection enables proactive maintenance to prevent faults and failures.