Research on High-Precision Detection of Bird Nests on Transmission Lines Based on ER-YOLO Algorithm

With the increasing awareness of nature conservation, more and more birds choose to build nests on high-voltage transmission lines, which may damage the insulation performance of transmission towers and cause power system failures, thereby seriously affecting the stability and safety of the power grid. In order to solve the problems of laborintensive, high cost, and high risk in traditional manual inspection, using drones equipped with automatic inspection software to detect bird nests on transmission lines has gradually become one of the mainstream inspection methods. This paper proposes an ER-YOLO algorithm based on the YOLOv10 model. The introduction of ECA attention mechanism enhances the small target feature extraction capability of YOLOv10 model. In addition, by using the Repulsion loss function, the interference of background objects on the detection box is reduced, the detection accuracy is improved, and the generalization ability of the model is enhanced through data augmentation. The experimental results show that the average precision of the ER-YOLO algorithm in bird nest detection task reaches 98.9%, which is an improvement compared to YOLOv10. Finally, it was verified through ablation experiments that the ECA module and data augmentation effectively improved detection performance. The research results of this paper show that the ER-YOLO algorithm has extremely superior small object detection ability in complex backgrounds.