A Method of Insulator Faults Detection in Aerial Images for High-Voltage Transmission Lines Inspection

• DOI: 10.3390/app9102009
• Publication Date: 2019-05-16T15:21:22Z
• Authors: Jiaming Han; Zhong Yang; Qiuyan Zhang; Cong Chen; Hongchen Li; Shangxiang Lai; Guoxiong Hu; Changliang Xu; Hao Xu; Di Wang; Rui Chen
• Keywords: Aerial image; Robustness

Insulator faults detection is an important task for high-voltage transmission line inspection. However, current methods often suffer from the lack of accuracy and robustness. Moreover, these can only detect one fault in insulator string, but cannot a multi-fault. In this paper, novel method proposed multi-fault UAV-based aerial images, backgrounds which usually contain much complex interference. The shapes insulators also vary obviously due to changes filming angle distance. To reduce impact interference on detection, we make full use deep neural network distinguish between background First all, plenty images with manually labelled ground-truth are collected construct standard dataset ‘InST_detection’. Secondly, new convolutional obtain accurate string positions image. Finally, that both images. Experimental results large number show our more effective efficient than state-of-the-art methods.