An enhanced algorithm for cell-level anomaly segmentation in photovoltaic solar panels using electroluminescence imaging

Ruiyao Duan, Yongjian Wang, Xisong Chen and Shihua Li

Energy, 2025, vol. 331, issue C

Abstract: Electroluminescence imaging is a crucial diagnostic tool for assessing the quality and performance of photovoltaic (PV) modules. However, current research often focuses on defect detection in PV modules, neglecting the need for detailed segmentation at the individual cell level. To address the limitations of existing methods in recognizing complex defects and suppressing background noise, this paper proposes a novel semantic segmentation algorithm (CAAK-Net), capable of identifying anomalies at the cell level of photovoltaic panels. To enhance network performance, CAAK-Net uses K-Net as a baseline and incorporates the Convolutional Block Attention Module (CBAM), Attention Refinement Module (ARM), and Atrous Spatial Pyramid Pooling (ASPP) modules. Experimental results comparing CAAK-Net with mainstream segmentation networks demonstrate its superior segmentation performance, particularly in recognizing defect edges and small-area attached defects. Additionally, we establish a quantitative correlation model between detection errors and PV system efficiency losses, pioneering the translation of pixel-level segmentation accuracy into measurable operational costs, thereby providing an economic assessment framework for industrial PV inspection. Furthermore, the network exhibits a certain degree of robustness in noisy environments, showcasing its segmentation advantages in diverse defect scenarios.

Keywords: Photovoltaic solar panels; Electroluminescence imaging; Semantic segmentation; Defect identification (search for similar items in EconPapers)
Date: 2025
References: Add references at CitEc
Citations:

Downloads: (external link)
http://www.sciencedirect.com/science/article/pii/S0360544225023539
Full text for ScienceDirect subscribers only

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:eee:energy:v:331:y:2025:i:c:s0360544225023539

DOI: 10.1016/j.energy.2025.136711

Access Statistics for this article

Energy is currently edited by Henrik Lund and Mark J. Kaiser

More articles in Energy from Elsevier
Bibliographic data for series maintained by Catherine Liu ().