Detection Algorithm of Untwisted or Broken Strands of Transmission Line Based on FRAC
Mingxian Guo, Jinbo Fang, Jun Tang, Tan Shou-biao · 2019 3rd International Conference on Electronic Information Technology and Computer Engineering (EITCE) · 2019
The detection of untwisted or broken strands of the transmission line plays an important role in ensuring the longterm and steady operation of the transmission system. This paper proposes an enhanced Faster R-CNN with Attention Channel (FRAC) and a sliding crop algorithm and use them in an well-designed algorithm flow suitable for the detection of the strands defects. Firstly, the image is merged with an attention map produced by attention network and become a 4-channel picture. Then the sliding crop algorithm cuts it into subimages before sending it into our proposed FRAC. The experiment results demonstrate that this algorithm flow has a significant improvement over the original Faster R-CNN network.