RCNN-based foreign object detection for securing power transmission lines (RCNN4SPTL)

• DOI: 10.1016/j.procs.2019.01.232
• Publication Date: 2019-02-10T15:19:58Z
• Authors: Weishan Zhang; Xia Liu; Jiangru Yuan; Liang Xu; Haoyun Sun; Jiehan Zhou; Xin Liu
• Keywords: RCNN; Pattern recognition; 0202 electrical engineering, electronic engineering, information engineering; 02 engineering and technology; Power transmission line; CNN

Abstract This paper proposes a new deep learning network - RCNN4SPTL (RCNN -based Foreign Object Detection for Securing Power Transmission lines), which is suitable for detecting foreign objects on power transmission lines. The RCNN4SPTL uses RPN (Region Proposal Network) to generate aspect ratio of the region proposals to align with the size of foreign objects. The RCNN4SPTL uses an end to end training to improve its performance. Experimental results show that the RCNN4SPTL significantly improves the detection speed and recognition accuracy, compared with the original Faster RCNN.