IRMA-International.org: Creator of Knowledge
Information Resources Management Association
Advancing the Concepts & Practices of Information Resources Management in Modern Organizations

Insulator Fault Detection From UAV Images Using YOLOv5

Insulator Fault Detection From UAV Images Using YOLOv5
View Sample PDF
Author(s): S. Venkata Suryanarayana (CVR College of Engineering, India), Katakam Koushik (CVR College of Engineering, India)and Prabu Sevugan (Pondicherry University, India)
Copyright: 2023
Pages: 13
Source title: Handbook of Research on Deep Learning Techniques for Cloud-Based Industrial IoT
Source Author(s)/Editor(s): P. Swarnalatha (Department of Information Security, School of Computer Science and Engineering, Vellore Institute of Technology, India)and S. Prabu (Department Banking Technology, Pondicherry University, India)
DOI: 10.4018/978-1-6684-8098-4.ch005

Purchase

View Insulator Fault Detection From UAV Images Using YOLOv5 on the publisher's website for pricing and purchasing information.

Abstract

Identification of insulator defects is one of the most important goals of an intelligent examination of high-voltage transmission lines. Because they provide mechanical support for electric transmission lines as well as electrical insulation, insulators are essential to the secure and reliable operation of power networks. A fresh dataset is first built by collecting aerial pictures in various scenes that have one or more defects. A feature pyramid network and an enhanced loss function are used by the CSPD-YOLO model to increase the precision of insulator failure detection. The insulator defective data set, which has two classes (insulator, defect), is used by the suggested technique to train and test the model using the YOLOv5 object detection algorithm. The authors evaluate how well the YOLOv3, YOLOv5, and related families perform when trained on the insulator defective dataset. Practitioners can use this information to choose the appropriate technique based on the insulator defective dataset.

Related Content

Fatima Ahmed Mohamed Abdalla, Noor Asiah Rashid. © 2026. 32 pages.
Fatima Ahmed Mohamed Abdalla, Noor Asiah Rashid. © 2026. 32 pages.
Azana Hafizah Mohd Aman, Wan Muhd Hazwan Azamuddin, Maznifah Salam, Zainab S. Attarbashi. © 2026. 32 pages.
Azana Hafizah Mohd Aman, Wan Muhd Hazwan Azamuddin, Maznifah Salam, Zainab S. Attarbashi. © 2026. 36 pages.
Salaheldin Mohamed Ibrahim Edam. © 2026. 42 pages.
Rubi Kadyan, Sunita Rani, Vinod Kr. Saroha. © 2026. 46 pages.
Mamoon M. Saeed, Zeinab E. Ahmed, Rania A. Mokhtar, Rashid A. Saeed. © 2026. 34 pages.
Body Bottom