Open AccessDetection of Common Foreign Objects on Power Grid Lines Based on Faster R-CNN Algorithm and Data Augmentation Method
Author(s) -
Guangxin Zu,
Guorong Wu,
Chong Zhang,
Xing He
Publication year - 2021
Publication title -
journal of physics. conference series
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.21
H-Index - 85
eISSN - 1742-6596
pISSN - 1742-6588
DOI - 10.1088/1742-6596/1746/1/012039
Subject(s) - computer science , algorithm , brightness , recall rate , grid , power grid , data set , artificial intelligence , training set , object detection , set (abstract data type) , computer vision , power (physics) , pattern recognition (psychology) , mathematics , physics , geometry , quantum mechanics , optics , programming language
Foreign objects hanging on the electrical equipment are big threats to the safety of the power grid. In this paper, a high-precision object detection model designed for common foreign objects in power grid based on Faster R-CNN algorithm is proposed. The data augmentation methods such as rotation, adjusting brightness and colour saturation, adding Gaussian noise are carried out to expand the data in the training set. The accuracy rate and the recall rate of the model reach 93.09% and 94.33% respectively. The object detection model trained by the training set with data augmentation has achieved better detection results.