Multimodal Segmentation for Photovoltaic Module Defect Detection
Open Access
- 5 December 2025
- journal article
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 13, 208652-208665
Abstract
Accurate photovoltaic module defect detection is hindered by limited single-modality information and weak feature representation under complex environmental conditions. To enhance defect identification in photovoltaic modules, a cross-modal attention and semantic-aware segmentation network is proposed herein. The framework introduces a cross-modal attention interaction module that adaptively balances visible and thermal features to strengthen inter-modal complementarity. A semantic-aware fusion module is also developed to refine semantic representation and improve feature fusion consistency. In addition, a dedicated multimodal dataset is constructed to support model evaluation. The proposed method demonstrates superior robustness and generalization capability across diverse scenarios. This work provides an effective and scalable approach for improving defect detection performance in photovoltaic systems via multimodal feature integration and attention-guided semantic learning.Keywords
Funding Information
- Graduate Student Scientific Research and Innovation Project of Hubei Minzu University (MYK2025075)
- Hubei Provincial Central Government-Guided Local Science and Technology Development Special Project (2022BGE235)
- Hubei Provincial Natural Science Foundation of China (2025AFD161)
- Hubei Provincial Natural Science Foundation of China (2023AFD061)
- Outstanding Youth Science and Technology Innovation Team Project for Colleges and Universities of Hubei Province of China (T2023013)
This publication has 45 references indexed in Scilit:
- Impacts of abatement in anthropogenic emissions in the context of China’s carbon neutrality on global photovoltaic potentialnpj Climate and Atmospheric Science, 2025
- Semantic segmentation method of photovoltaic cell microcracks based on EL polarization imagingSolar Energy, 2025
- Channel2DTransformer: A Multi-level Features Self-attention Fusion Module for Semantic SegmentationInternational Journal of Computational Intelligence Systems, 2024
- SIIF: Semantic information interactive fusion network for photovoltaic defect segmentationApplied Energy, 2024
- ECFNet: Efficient cross-layer fusion network for real time RGB-Thermal urban scene parsingDigital Signal Processing, 2024
- Hotspot defect detection for photovoltaic modules under complex backgroundsMultimedia Systems, 2023
- GCNet: Grid-like context-aware network for RGB-thermal semantic segmentationNeurocomputing, 2022
- A novel framework on intelligent detection for module defects of PV plant combining the visible and infrared imagesSolar Energy, 2022
- Attention fusion network for multi-spectral semantic segmentationPattern Recognition Letters, 2021
- Technology, cost, economic performance of distributed photovoltaic industry in ChinaRenewable and Sustainable Energy Reviews, 2019