A Defect Detection Method Based on Faster RCNN for Power Equipment. Issue 1 (February 2021)

• Record Type: Journal Article
• Authors: Cheng, Yang; Xia, Lingzhi; Yan, Bo; Chen, Jiang; Hu, Dongsheng; Zhu, Lvfu
• Is Part Of: Journal of physics. Volume 1754:Issue 1(2021)
• Publication Date: 2021-02
• Subjects: Physics -- Congresses; 530.5
• DOI: 10.1088/1742-6596/1754/1/012025
• Languages: English
• ISSNs: 1742-6588

Abstract

Abstract: With the wide application of infrared image acquisition technology in power inspection, a large number of infrared images of power equipment have been obtained. The traditional machine learning method has low accuracy and poor generalization. Therefore, in this paper, the deep learning technology is applied to infrared image detection of power equipment, and a defect detection method based on Faster region convolution neural network (RCNN) is proposed. In this method, the deep residual network is used to extract image features, and the regional proposal network is optimized according to the shape characteristics of power equipment, and the network is trained with the help of shared convolution layer. The experimental results show that the proposed method has high detection accuracy, good robustness and generalization ability.