A Scalable Framework for Annotating Photovoltaic Cell Defects in Electroluminescence Images
Autores: Iñigo Martínez López
Fecha: 01.09.2023
IEEE Transactions on Industrial Informatics
Abstract
The correct functioning of photovoltaic (PV) cells is critical to ensuring the optimal performance of a solar plant. Anomaly detection techniques for PV cells can result in significant cost savings in operation and maintenance (O&M). Recent research has focused on deep learning techniques for automatically detecting anomalies in electroluminescence images. Automated anomaly annotations can improve current O&M methodologies and help develop decision-making systems to extend the life cycle of the PV cells and predict failures. This article addresses the lack of anomaly segmentation annotations in the literature by proposing a combination of state-of-the-art data-driven techniques to create a golden standard benchmark. The proposed method stands out for: 1) its adaptability to new PV cell types; 2) cost-efficient fine-tuning; and 3) leverage public datasets to generate advanced annotations. The methodology has been validated in the annotation of a widely used dataset, obtaining a reduction of the annotation cost by 60%.
BIB_text
title = {A Scalable Framework for Annotating Photovoltaic Cell Defects in Electroluminescence Images},
journal = {IEEE Transactions on Industrial Informatics},
pages = {9361-9369},
volume = {19},
keywds = {
Anomaly segmentation; benchmark dataset; deep learning; defect annotation; electroluminescence (EL); golden standard; photovoltaic (PV) cells
}
abstract = {
The correct functioning of photovoltaic (PV) cells is critical to ensuring the optimal performance of a solar plant. Anomaly detection techniques for PV cells can result in significant cost savings in operation and maintenance (O&M). Recent research has focused on deep learning techniques for automatically detecting anomalies in electroluminescence images. Automated anomaly annotations can improve current O&M methodologies and help develop decision-making systems to extend the life cycle of the PV cells and predict failures. This article addresses the lack of anomaly segmentation annotations in the literature by proposing a combination of state-of-the-art data-driven techniques to create a golden standard benchmark. The proposed method stands out for: 1) its adaptability to new PV cell types; 2) cost-efficient fine-tuning; and 3) leverage public datasets to generate advanced annotations. The methodology has been validated in the annotation of a widely used dataset, obtaining a reduction of the annotation cost by 60%.
}
doi = {10.1109/TII.2022.3228680},
date = {2023-09-01},
}
