Recognition of main electrical equipment images based on YOLOv3
Recognition of main electrical equipment images based on YOLOv3
- Author(s): L. Niu 1 ; S. Wang 1 ; F. Lü 1
- DOI: 10.1049/icp.2020.0049
For access to this article, please select a purchase option:
Buy conference paper PDF
Buy Knowledge Pack
IET members benefit from discounts to all IET publications and free access to E&T Magazine. If you are an IET member, log in to your account and the discounts will automatically be applied.
Thank you
Your recommendation has been sent to your librarian.
- Author(s): L. Niu 1 ; S. Wang 1 ; F. Lü 1
-
-
View affiliations
-
Affiliations:
1:
North China Electric Power University School of Electrical and Electronic Engineering , Beijing , China
Source:
The 16th IET International Conference on AC and DC Power Transmission (ACDC 2020),
2021
p.
1350 – 1353
-
Affiliations:
1:
North China Electric Power University School of Electrical and Electronic Engineering , Beijing , China
- Conference: The 16th IET International Conference on AC and DC Power Transmission (ACDC 2020)
- DOI: 10.1049/icp.2020.0049
- ISBN: 978-1-83953-330-3
- Location: Online Conference
- Conference date: 02-03 July 2020
- Format: PDF
With the advantages of high sensitivity, non-contact and intuitive characterization in discharge detection, the solar-blind UV imaging detection method has been used in the UAV and robotic inspections. Based on the characteristics of dual-channel imaging of UV imager, this paper proposes a visible channel image insulator recognition method based on YOLOv3.The hardware and software platform of deep learning based on the TensorFlow and Darknet framework in the Linux environment was built. The labeled database of the main electrical equipment was established and then the training of the convolution neural network was complete. The influence of the number of training samples and the layers of the network on the recognition accuracy was studied. The optimization method of the network parameters and the super parameter was given. With the network optimized, the recognition accuracy on the test set increasing by 3% reached 95%. This paper may provide guidance for engineering applications.
Inspec keywords: convolutional neural nets; image recognition; power engineering computing; deep learning (artificial intelligence); power apparatus; feature extraction; Linux
Subjects: Neural nets; Power engineering computing; Computer vision and image processing techniques; Image recognition; Operating systems