Based on the MFBN-YOLOv5 transmission line insulator defect detection research

In order to solve the problem of low efficiency of insulator detection on transmission lines, an insulator defect detection model MFBN-YOLOv5 is proposed in this paper. Firstly, to solve the problem of insufficient feature extraction capability in the backbone network, MC3 module is designed to replace C3 module in the backbone network, and Fenhence module is introduced in the tail of the backbone network. The module adopts convolution and void convolution series to effectively expand the receptive field and enhance feature extraction. Secondly, in order to improve the feature fusion ability of the model, BiFPN structure is introduced in the neck to improve the feature fusion ability of the network for different defects of insulators. The experimental results show that the improved MFBN-YOLOv5 model can detect insulator defects quickly and accurately, with an average accuracy of 95.6% (mAP 0.5), which is 3.9% higher than that of the original YOLOv5s model, and can meet the needs of daily power inspection.