Research on insulator recognition algorithm based on deep learning

Unmanned aerial vehicle (UAV) was gradually replacing manual inspection of transmission lines to detect the state of insulators. In this paper, based on the image data source collected by UAV inspection, a method of insulator feature recognition and location by multi-scale grid was established based...

Full description

Saved in:
Bibliographic Details
Published in2022 International Conference on Cloud Computing, Big Data Applications and Software Engineering (CBASE) pp. 237 - 241
Main Authors Kuang, Fan, Cui, Jianhua, Dai, Xiliang, Zhang, Cheng
Format Conference Proceeding
LanguageEnglish
Published IEEE 01.09.2022
Subjects
Autonomous aerial vehicles
Deep learning
identification
Inspection
insulator
Insulators
Soft sensors
Training
Transmission line measurements
Online AccessGet full text

Cover

More Information
Summary:Unmanned aerial vehicle (UAV) was gradually replacing manual inspection of transmission lines to detect the state of insulators. In this paper, based on the image data source collected by UAV inspection, a method of insulator feature recognition and location by multi-scale grid was established based on deep learning theory. Under the structure of multi-scale grid recognition algorithm, feature extraction module, scale scaling module, classification and regression module were designed. After training and testing of the image data, the results of insulator identification and location were obtained. And the results showed that this method could effectively identify the characteristics of insulators.
DOI:10.1109/CBASE57816.2022.00052