Detection of Fault Insulator of Power Transmission Line Based on Region-CNN

Abstract

Insulators are important electrical components to ensure the normal and reliable operation of power transmission lines. It has become an indispensable task to detect faulty insulators in actual production processes successfully. In the current big data environment, the time and labor cost of traditional detection methods is every high. In this paper, we propose an smart detection method of power transmission live based on R-CNN, which use the convolutional neural network (CNN) to extract visual features from the aerial images. Our method can detects the area of the image where the insulator is located and then perform secondary detection on the basis of the entire insulator. We use drone aerial images as experimental data to verify the identification of insulator self-explosion defects. The experimental results show that our method can accurately detect insulators in different environments and accurately detect the faulty insulators in the image. We can confirm our method is of great robustness and practicality in insulator detection.