An Improved Faster R-CNN Algorithm for Electric Equipment Detection
Jianlong Guo, Weixia Feng, Manhua Wen · 2019
More and more images and videos have been collected in power system, however, it is still a challenge to recognize the electric equipment in the unstructured multimedia stream automatically. In this paper, we propose to apply an improved Faster R-CNN algorithm to detect and recognize the electric equipment for the electric equipment images. To be more specific, we adjust the key parameters of the Faster R-CNN framework to detect and recognize the electric equipment simultaneously. In order to improve the recognition accuracy, we propose DisturbIoU algorithm to prevent the network training from over-fitting. Experiment results show that for insulator, transformer and power pole, the average recognition accuracy rate of the proposed algorithm is 96.33%, 99.89%, 97.98% respectively, which is 3.28%, 6.02% and 6.21% higher than that of the traditional Faster R-CNN algorithm.