Insulator defect identification for transmission line drone inspection based on YOLOv8-transformer

The insulator image data collected by UAVs are easily affected by environmental factors (such as illumination and angle), and it is difficult to reflect the global characteristics of insulator defects, resulting in insufficient recognition accuracy. Therefore, a method for identifying insulator defects in transmission line UAV inspection based on YOLOv8- Transformer is studied. The insulator defects of small targets are normalized, and grayscale adjustment is used to reduce the influence of illumination and retain color features. The YOLOv8 algorithm is used to locate the insulator area in the UAV inspection image data. Transformer is used to capture the global information near the insulator, and the position information and global features are fused and input into the logistic regression classifier to determine whether the insulator has defects and obtain the type of insulator defects in transmission line UAV inspection. The experimental results show that this method has clear boundaries and higher recognition efficiency in the identification of insulator defects in mountainous transmission line UAV inspection and performs best in the confusion matrix of insulator defect classification and recognition, with good overall recognition accuracy and good application prospects.