Photovoltaic Module Infrared Defect Image Generation Using Coordinate Awareness and Subpixel Reconstruction

Xujiang Liu; Guangyu Zhu; Ziwen Song; Tingting Yang; Dongjing Li; He Wang

doi:null

您当前的位置：

首页 >

文章列表页 >

Photovoltaic Module Infrared Defect Image Generation Using Coordinate Awareness and Subpixel Reconstruction

更新时间：2026-09-02

- Photovoltaic Module Infrared Defect Image Generation Using Coordinate Awareness and Subpixel Reconstruction
- JOURNAL OF XI’AN JIAOTONG UNIVERSITY (2026)
- 作者机构：
  
  1.西安交通大学能源与动力工程学院，710049，陕西西安
  2.内蒙古智慧运维新能源有限公司，010010，内蒙古呼和浩特
- 作者简介：
- 基金信息：
- DOI：
  CLC： TP183;TK519
- Received：09 March 2026，
  
  Revised：2026-07-06，
  
  Accepted：24 July 2026，
- 稿件说明：
移动端阅览
Xujiang Liu, Guangyu Zhu, Ziwen Song, et al. Photovoltaic Module Infrared Defect Image Generation Using Coordinate Awareness and Subpixel Reconstruction[J]. JOURNAL OF XI’AN JIAOTONG UNIVERSITY,2026.
DOI：

Xujiang Liu, Guangyu Zhu, Ziwen Song, et al. Photovoltaic Module Infrared Defect Image Generation Using Coordinate Awareness and Subpixel Reconstruction[J]. JOURNAL OF XI’AN JIAOTONG UNIVERSITY,2026. DOI：

摘要

针对光伏红外图像缺陷识别任务中小样本和类别分布不均衡导致的诊断模型泛化能力不足问题，构建了一种采用坐标感知与亚像素重构的生成对抗网络（CPS-GAN），并以其为核心提出面向小样本类别数据扩充的光伏组件红外缺陷图像生成方法。首先，针对光伏组件红外图像局部热异常位置敏感、热梯度变化明显等特征，在生成器中引入类别条件约束与CoordConv坐标感知模块，建立潜在编码与几何空间之间的映射关系，改善传统GAN难以准确表征局部缺陷空间位置的问题。其次，采用基于PixelShuffle的亚像素重构方法替代传统转置卷积，抑制生成图像中的“棋盘伪影”，并保留热异常分布特征。最后，结合PatchGAN判别器强化局部纹理判别。实验结果表明：CPS-GAN在多个定量评价指标及语义评价方案上均优于DCGAN与WGAN-GP基线模型，并在t-SNE特征可视化中表现出更好的特征分布对齐效果。在无需分类器架构优化的条件下，基于CPS-GAN增强数据集训练的分类器获得了84.00%的测试准确率，较原始数据提升了7.05%。所提方法能够有效生成高质量、多样化的光伏组件红外缺陷样本，为提升光伏智能诊断系统的鲁棒性提供数据支撑。

Abstract

To address the limited generalization capability of diagnostic models caused by small sample sizes and class imbalance in photovoltaic infrared defect recognition

a generative adversarial network incorporating coordinate awareness and subpixel reconstruction

termed CPS-GAN

is developed. With CPS-GAN as the core generative network

a photovoltaic module infrared defect image generation method is further proposed for augmenting defect classes with limited samples. First

considering the spatial sensitivity of local thermal anomalies and the pronounced variations in thermal gradients in photovoltaic infrared images

class-conditional constraints and a CoordConv-based coordinate-aware module are introduced into the generator. This design establishes a mapping relationship between latent representations and geometric space

thereby improving the ability of conventional GANs to represent the spatial locations of local defects accurately. Second

a PixelShuffle-based subpixel reconstruction strategy is employed to replace conventional transposed convolution

suppress checkerboard artifacts in generated images

and preserve thermal anomaly distribution characteristics. Finally

a PatchGAN discriminator is incorporated to strengthen local texture discrimination. Experimental results demonstrate that CPS-GAN outperforms the DCGAN and WGAN-GP baselines across multiple quantitative metrics and semantic evaluation protocols

while exhibiting improved feature-distribution alignment in t-SNE visualizations. Without modifying the classifier architecture

the classifier trained on the CPS-GAN-augmented dataset achieves a test accuracy of 84.00%

representing an improvement of 7.05 percentage points over training with the original dataset. The proposed method can effectively generate high-quality and diverse infrared defect samples of photovoltaic modules

thereby providing reliable data support for improving the robustness of intelligent photovoltaic diagnostic systems.

关键词

Keywords

references

别朝红 , 潘超琼 , 陈叶 , 等 . 能源转型下新能源电力系统概率风险评估 [J ] . 西安交通大学学报 , 2021 , 55 ( 7 ): 1 - 11 .

Bie Zhaohong , Pan Chaoqiong , Chen Ye , et al . Probabilistic risk assessment of new energy power system in the context of energy transition: a review [J ] . Journal of Xi'an Jiaotong University , 2021 , 55 ( 7 ): 1 - 11 .

Liu Xujiang , Zhu Guangyu , Li Zhigang , et al . A comprehensive evaluation framework using generative adversarial networks for infrared defect images in photovoltaic modules [J ] . Solar Energy , 2026 , 306 : 114311 .

Chen Xin , Karin T , Jain A . Automated defect identification in electroluminescence images of solar modules [J ] . Solar Energy , 2022 , 242 : 20 - 29 .

Et-Taleby A , Chaibi Y , Benslimane M , et al . Applications of machine learning algorithms for photovoltaic fault detection: a review [J ] . Statistics , Optimization & Information Computing, 2023 , 11 ( 1 ): 168 - 177 .

Venkatakrishnan G R , Rengaraj R , Tamilselvi S , et al . Detection, location, and diagnosis of different faults in large solar PV system—a review [J ] . International Journal of Low-Carbon Technologies , 2023 , 18 : 659 - 674 .

Kellil N , Aissat A , Mellit A . Fault diagnosis of photovoltaic modules using deep neural networks and infrared images under Algerian climatic conditions [J ] . Energy , 2023 , 263, Part C: 125902.

Zhao Shenshen , Chen Haiyong , Wang Chuhan , et al . SNCF-Net: scale-aware neighborhood correlation feature network for hotspot defect detection of photovoltaic farms [J ] . Measurement , 2023 , 206 : 112342 .

翟金星 , 王力平 , 张艳辉 . 基于无人机红外视觉的光伏电站热斑智能巡检 [J ] . 红外技术 , 2026 , 48 ( 1 ): 95 - 104 .

Zhai Jinxing , Wang Liping , Zhang Yanhui . Unmanned-Aerial-Vehicle Infrared-Vision-Based Intelligent Inspection of Hotspots in Photovoltaic Power Station [J ] . Infrared Technology , 2026 , 48 ( 1 ): 95 - 104

陈登峰 , 刘茜茜 , 刘世鹏 , 等 . 面向光伏组件缺陷检测的多源图像融合方法研究 [J ] . 太阳能学报 , 2026 , 47 ( 4 ): 622 - 629 .

Chen Dengfeng , Liu Qianqian , Liu Shipeng , et al . Multi-source image fusion methods for photovoltaic module defect detection [J ] . Acta Energiae Solaris Sinica , 2026 , 47 ( 4 ): 622 - 629 .

项新建 , 孙思琪 , 罗秋霞 , 等 . 基于跨层次局部特征融合的光伏热斑缺陷检测 [J ] . 太阳能学报 , 2026 , 47 ( 3 ): 732 - 738 .

Xiang Xinjian , Sun Siqi , Luo Qiuxia , et al . Photovoltaic Hotspot Defect Detection Based on Cross-Level Local Feature Fusion [J ] . Acta Energiae Solaris Sinica , 2026 , 47 ( 3 ): 732 - 738 .

Le M , Luong V S , Nguyen D K , et al . Remote anomaly detection and classification of solar photovoltaic modules based on deep neural network [J ] . Sustainable Energy Technologies and Assessments , 2021 , 48 : 101545 .

Duranay Z B . Fault detection in solar energy systems: a deep learning approach [J ] . Electronics , 2023 , 12 ( 21 ): 4397 .

Korkmaz D , Acikgoz H . An efficient fault classification method in solar photovoltaic modules using transfer learning and multi-scale convolutional neural network [J ] . Engineering Applications of Artificial Intelligence , 2022 , 113 : 104959 .

孙廨尧 , 孙俊海 , 朱先远 , 等 . 基于改进YOLOv8的红外光伏组件缺陷检测方法 [J ] . 太阳能学报 , 2026 , 47 ( 2 ): 148 - 184 .

Sun Xieyao , Sun Junhai , Zhu Xianyuan , et al . Infrared Photovoltaic Module defect Detection Method Based on Improved YOLOv8 [J ] . Acta Energiae Solaris Sinica , 2026 , 47 ( 2 ): 148 - 184 .

Sinap V , Kumtepe A . CNN-based automatic detection of photovoltaic solar module anomalies in infrared images: a comparative study [J ] . Neural Computing and Applications , 2024 , 36 ( 28 ): 17715 - 17736 .

Le M , Le D V , Ha Thi Vu H . Thermal inspection of photovoltaic modules with deep convolutional neural networks on edge devices in AUV [J ] . Measurement , 2023 , 218 : 113135 .

Tang Cheng , Ren Hui , Xia Jing , et al . Automatic defect identification of PV panels with IR images through unmanned aircraft [J ] . IET Renewable Power Generation , 2023 , 17 ( 12 ): 3108 - 3119 .

Ho J , Jain A , Abbeel P . Denoising diffusion probabilistic models [C ] // Proceedings of the 34th International Conference on Neural Information Processing Systems . Red Hook, NY, USA : Curran Associates Inc. , 2020 : 6840 - 6851 .

Goodfellow I J , Pouget-Abadie J , Mirza M , et al . Generative adversarial nets [C ] // Proceedings of the 28th International Conference on Neural Information Processing Systems . Cambridge, MA, USA : MIT Press , 2014 : 2672 - 2680 .

Heusel M , Ramsauer H , Unterthiner T , et al . GANs trained by a two time-scale update rule converge to a local nash equilibrium [C ] // Proceedings of the 31st International Conference on Neural Information Processing Systems . Red Hook, NY, USA : Curran Associates Inc. , 2017 : 6629 - 6640 .

Gretton A , Borgwardt K M , Rasch M , et al . A kernel method for the two-sample-problem [C ] // Proceedings of the 20th International Conference on Neural Information Processing Systems . Cambridge, MA, USA : MIT Press , 2006 : 513 - 520 .

Wang Z , Simoncelli E P , Bovik A C . Multiscale structural similarity for image quality assessment [C ] // The Thrity-Seventh Asilomar Conference on Signals , Systems & Computers, 2003 . Piscataway, NJ, USA : IEEE , 2003 : 1398 - 1402 .

Zhang R , Isola P , Efros A A , et al . The unreasonable effectiveness of deep features as a perceptual metric [C ] // 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition . Piscataway, NJ, USA : IEEE , 2018 : 586 - 595 .

Shmelkov K , Schmid C , Alahari K . How good is my GAN? [C ] // Computer Vision – ECCV 2018 . Cham : Springer International Publishing , 2018 : 218 - 234 .

Radford A , Metz L , Chintala S . Unsupervised representation learning with deep convolutional generative adversarial networks [C ] // Proceedings of the 4th International Conference on Learning Representations (ICLR 2016) . San Juan, Puerto Rico : OpenReview.net , 2016 . arXiv: 1511 . 06434

Gulrajani I , Ahmed F , Arjovsky M , et al . Improved training of Wasserstein GANs [C ] // Proceedings of the 31st International Conference on Neural Information Processing Systems . Red Hook, NY, USA : Curran Associates Inc. , 2017 : 5769 - 5779 .

Tsinganos P , Cornelis B , Cornelis J , et al . Data augmentation of surface electromyography for hand gesture recognition [J ] . Sensors , 2020 , 20 ( 17 ): 4892 .

闫号 , 戴佳佳 , 龚小溪 , 等 . 基于多源图像融合的光伏面板缺陷检测 [J ] . 红外技术 , 2023 , 45 ( 5 ): 488 - 497 .

Yan Hao , Dai Jiajia , Gong Xiaoxi , et al . Defect Detection of Photovoltaic Panel Based on Multisource Image Fusion [J ] . Infrared Technology , 2023 , 45 ( 5 ): 488 - 497 .

Dong Chunru , Chen Linlin , Liu Yuchen , et al . An improved feature aggregation network for photovoltaic cell defect detection [J ] . Journal of Renewable and Sustainable Energy , 2025 , 17 ( 1 ): 013503 .

谢雨杰 , 杜友田，张潇 , 等 . 面向新概念学习的图像描述生成模型 [J ] . 西安交通大学学报 , 2020 , 54 ( 12 ): 37 - 44 .

Xie Yujie , Du Youyu , Zhang Xiao , et al . An Image Description Generation Model for the Learning of Novel Concepts [J ] . Journal of Xi'an Jiaotong University , 2020 , 54 ( 12 ): 37 - 44 .

Boubaker S , Kamel S , Ghazouani N , et al . Assessment of machine and deep learning approaches for fault diagnosis in photovoltaic systems using infrared thermography [J ] . Remote Sensing , 2023 , 15 ( 6 ): 1686 .

Wang B , Chen Q , Wang M , et al . Photovoltaic fault classification dataset (PVF-10): A high-resolution unmanned aerial vehicle thermal infrared image dataset for fine-grained photovoltaic fault classification [J ] . Applied Energy , 2024 , 376 : 124187 .

Views

下载量

CSCD

Alert me when the article has been cited

提交

Tools

Publicity Resources

Interpreting the Invariance Mechanism of Data Augmentation on Deep Convolutional Networks

Interpretable Mechanical Fault Diagnosis Method Based on Dual-Tree Complex Wavelet Transform

Segment Anything Model Guided Weakly-Supervised Mutual Learning for 3D Medical Image Registration

Phase-Field Analysis Method for the Reliability of High-g Micro-Electro-Mechanical System Accelerometers

Progress in Visualization Experimental Studies on Heat Transfer of Supercritical Fluids

Related Author

LIU Ni

LEI Jiao Jiao

XING ZHEN RONG

HAN WEN JING

HE LI HUO

Related Institution

School of Information Engineering, Chang’an University

School of Electronic Engineering, Xidian University

⁰