Oct 2026· International Conference on Image Processing, Object Detection and Tracking· Vol 14506, pp. 1450614 - 1450614-7· 0 citations· 15 references
Engineering
TL;DR
Experimental results demonstrate that the proposed MBCSE module improves multi-scale defect representation while maintaining a lightweight network structure.
Abstract
Hidden defects in photovoltaic (PV) modules reduce power generation efficiency and may cause safety risks in large-scale PV plants. Existing deep learning detectors still suffer from insufficient perception of small-scale defects and high computational cost in deployment-oriented inspection. To address this problem with a simple and lightweight design, this paper proposes MBCSE-YOLO, a YOLOv11n-based defect detection model enhanced by a Multi-Branch Channel and Spatial Enhancement (MBCSE) module. The proposed MBCSE module uses parallel 1x1, 3x3, and 5x5 convolutional branches to extract defect features at different receptive fields, and then employs channel re-calibration to suppress redundant background information. Experiments on the Thermal Solar PV and PVEL-AD datasets show that MBCSE-YOLO improves mAP50 by 9.2% and 2.3%, respectively, compared with YOLOv11n. Meanwhile, mAP50:95 is improved by 8.5% and 2.0%, respectively, while GFLOPs and parameters are reduced by 6.1% and 1.2%. Experimental results demonstrate that the proposed MBCSE module improves multi-scale defect representation while maintaining a lightweight network structure.
Photovoltaic (PV) modules are widely used in building-integrated photovoltaic systems and renewable energy infrastructure. Accurate defect detection is essential for improving the reliability, energy efficiency, and sustainable operation of PV systems. However, existing deep learning-based inspection methods still su...
The increasing deployment of photovoltaic (PV) systems has created a growing demand for intelligent and reliable defect detection methods capable of supporting efficient maintenance and maximizing energy production. However, existing deep learning approaches often suffer from high computational complexity, sensitivity...
Zahraa Tarek, Hussein A. Kazem, Khaled Alnowaiser et al.· Emerging Science Journal· 0 citations
A lightweight spatial enhancement detection model, namely GLA-YOLO, based on YOLOv5s, GhostConv and C3Ghost are introduced to reduce computational complexity and parameter scale and to handle the small size, complex morphology and background interference of photovoltaic defects.
Hui-Jie Jia, Xiao-Hui Zhang, Hong-Biao Ma et al.· Engineering Research Express· 0 citations
Defect detection in photovoltaic (PV) panels is of great significance for ensuring power generation efficiency and the safe operation of power stations. To address the unstable performance of existing detection methods in identifying small defects, handling defect scale variations, and dealing with complex polycrystall...
Teng-Ming Gao, Si-Xian Wang· The International Conference...· 0 citations
Cell defects in photovoltaic (PV) modules directly affect production yield, making early and reliable detection critical. Electroluminescence (EL) imaging is widely used for this purpose since it reveals defects such as microcracks and inactive regions; however, manual assessment is time-consuming and subjective. In th...
Fatih Gündoğdu, M. Çalışan· Adıyaman Üniversitesi Mühend...· 0 citations
This study proposes a lightweight and enhanced You Only Look Once (YOLO)11-based architecture specifically designed for superior defect detection, which surpasses all competing methods by a substantial margin and is evaluated against a comprehensive suite of state-of-the-art object detection frameworks.
Mehmet Fatih Bekciogullari, H. Acikgoz, S. Ozbay· IEEE Access· 0 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.