This paper proposes a novel power transmission line monitoring method using unmanned aerial vehicles (UAVs) and YOLO-v8, a deep-learning object detection model. Traditional methods for detecting insulator defects, crucial for maintaining grid stability, are time-consuming and prone to human error. This challenge is addressed by leveraging the speed and accuracy of YOLO-v8 to automatically identify potential hazards in aerial images captured by UAVs. The YOLO-v8 model efficiently detected various insulator defects, including Damages and Flashover, by applying bounding boxes to relevant image regions. The trained model achieved a high overall accuracy of 94% (MAP50) on the dataset, with individual class accuracies reaching 97% (Healthy), 99% (Damage), and 85% (Flashover). Furthermore, data augmentation techniques further enhanced the classification accuracy to 98.6%.
