ORIGINAL RESEARCH article
Front. Artif. Intell.
Sec. Pattern Recognition
Volume 8 - 2025 | doi: 10.3389/frai.2025.1581010
This article is part of the Research TopicAdvances in Deep Learning for Perception Science: Modeling Mechanisms and ApplicationsView all articles
A Robust Corroded Metal Fitting Detection Approach for UAV intelligent Inspection with Knowledge-Distilled Lightweight YOLO Model
Provisionally accepted
- 1State Grid Henan Electric Power Research Institute, Zhengzhou, China
- 2Henan Normal University, Xinxiang, China
Select one of your emails
You have multiple emails registered with Frontiers:
Notify me on publication
Please enter your email address:
If you already have an account, please login
You don't have a Frontiers account ? You can register here
Despite the widespread application of unmanned aerial vehicles (UAVs) in daily transmission line inspection, detecting corroded metal fittings remains challenging due to two key issues: the small size of fittings and environmental interference in UAV-collected images, leading to severe false and missed detections; and the complexity of existing deep learning-based object detection models, which hinders edge device deployment and real-time performance, restricting the automation of smart grid operation and maintenance. To address these, this paper proposes a novel knowledge-distilled lightweight YOLO model, integrating a densely-connected convolutional network and spatial pixel-aware self-attention mechanism in the teacher model training stage to enhance feature transfer and structured feature utilization for reducing environmental interference, while employing the lightweight MobileNet as the feature extractor in the student model training stage and optimizing candidate box migration via the teacher model's efficient intersection over union non-maximum suppression (EIoU-NMS). This model combines theoretical innovation with practical value: it overcomes the challenges of small-object fitting detection in complex environments, improving fault identification accuracy and reducing manual inspection costs and missed detection risks, while its lightweight design enables rapid deployment and real-time detection on UAV terminals, providing a reliable technical solution for unmanned smart grid operation. Experimental results on actual UAV inspection images demonstrate that the model significantly enhances detection accuracy, reduces false and missed detections, achieves faster speeds with substantially fewer parameters, highlighting its outstanding effectiveness and practicality in power system maintenance scenarios.
Keywords: transmission line fitting, deep learning, object detection, Lightweight, Interpretability
Received: 25 Apr 2025; Accepted: 21 Aug 2025.
Copyright: © 2025 Tian, Zhang, Li, Liu and Mao. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
* Correspondence: Wentao Mao, Henan Normal University, Xinxiang, China
Disclaimer: All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article or claim that may be made by its manufacturer is not guaranteed or endorsed by the publisher.