An Accurate Detection Method for Insulator Dropout Based on Yolox Algorithm
Yuying Zhang, Shu Jia Yan, Qiang Chen, Rong Song, Mei Song Tong · 2023
In order to address the problem of insulator drops in aerial power line inspection tasks, this paper proposes a lightweight Yolox algorithm and conducts experiments to verify its effectiveness. The proposed algorithm can detect insulator strings and also small targets such as insulator drops. The Yolox-s model with a shallower depth is used for insulator drop detection, and other standard Yolox models are trained and compared to validate the algorithm. The experimental results show that the proposed model has the advantages of small size, reliable real-time performance, and high detection accuracy, which makes it more suitable for insulator drop detection than other models. The application of this algorithm is expected to improve the efficiency and accuracy of aerial power line inspection tasks and provide a guarantee for the safe operation of power transmission lines.