Solar Cell Defect Generation Algorithm Combining Multiple Perception Fields and Attention
ZHOU Ying1,2, PEI Shenghu1, CHEN Haiyong1,2, YAN Yuze1
1. School of Artificial Intelligence, Hebei University of Technology, Tianjin 300401; 2. Hebei Control Engineering Technology Research Center, Tianjin 300130
Abstract Aiming at the problem of insufficient image samples for some certain defects in solar cells, a solar cell defect generation algorithm combining multiple perception fields and attention is proposed. The generated images are utilized to train the defect detection model. Firstly, a generative adversarial network with dual discriminators is constructed, and a global discriminator and a local discriminator focuse on global information and local details, respectively. Secondly, the multiple perception field feature extraction is designed and fused with the improved attention module to form a multiple perception field attention module. The module is utilized in the network structure of both the generator and the discriminator. Finally, structural similarity loss and peak signal-to-noise ratio loss are added to the loss function for generator training, and the generated images are mean filtered. The generation experiments for 3 different scales of defect images on the solar electroluminescence dataset show that the structural similarity and peak signal-to-noise ratio are high. Additionally, after training the YOLOv7 detection model with the generated defect images, the average precision values for all three defects are high.
Fund:National Natural Science Foundation of China(No.62073117,U21A20482), Central Guidance on Local Science and Technology Development Funds(No.206Z1701G)
Corresponding Authors:
ZHOU Ying, Ph.D., associate professor. Her research interests include artificial intelligence and image processing.
About author:: PEI Shenghu, master student. His research interests include image processing.CHEN Haiyong, Ph.D., professor. His research interests include machine vision and image processing.YAN Yuze, master student. His research interests include target detection.
[1] 谢源,苗玉彬,许凤麟,等.基于半监督深度卷积生成对抗网络的注塑瓶表面缺陷检测模型.计算机科学, 2020, 47(7): 92-96. (XIE Y, MIAO Y B, XU F L, et al. Injection-Molded Bottle Defect Detection Using Semi-Supervised Deep Convolutional Generative Adversarial Network. Computer Science, 2020, 47(7): 92-96.) [2] BAEK F, KIM D, PARK S, et al. Conditional Generative Adversa-rial Networks with Adversarial Attack and Defense for Generative Data Augmentation. Journal of Computing in Civil Engineering, 2022, 36(3). DOI: 10.1061/(ASCE)CP.1943-5487.0001015 [3] GOODFELLOW I J, POUGET-ABADIE J, MIRZA M, et al. Generative Adversarial Nets // Proc of the 27th International Conference on Neural Information Proceeding Systems. Cambridge, USA: MIT Press, 2014: 2672-2680. [4] 张一平,许盛之,孟子尧,等.基于FC-ACGAN网络的太阳电池EL图像的数据增强方法.太阳能学报, 2021, 42(10): 35-41. (ZHANG Y P, XU S Z, MENG Z Y, et al. Data Enhancement Method of Solar Cell EL Images Based on FC-ACGAN Network. Acta Energiae Solaris Sinica, 2021, 42(10): 35-41.) [5] 刘坤,文熙,黄闽茗,等.基于生成对抗网络的太阳能电池缺陷增强方法.浙江大学学报(工学版), 2020, 54(4): 684-693. (LIU K, WEN X, HUANG M M, et al. Solar Cell Defect Enhancement Method Based on Generative Adversarial Network. Journal of Zhejiang University(Engineering Science), 2020, 54(4): 684-693.) [6] MA J Y, XU H, JIANG J J, et al. DDcGAN: A Dual-Discriminator Conditional Generative Adversarial Network for Multi-resolution Image Fusion. IEEE Transactions on Image Processing, 2020, 29: 4980-4995. [7] YANG X H, XU P, JIN H Y, et al. Semantic Face Completion Based on DCGAN with Dual-Discriminator // Proc of the 7th Annual International Conference on Network and Information Systems for Computers. Washington, USA: IEEE, 2021: 10-14. [8] TAN J X, LIAO X, LIU J T, et al. Channel Attention Image Ste-ganography with Generative Adversarial Networks. IEEE Transactions on Network Science and Engineering, 2022, 9(2): 888-903. [9] CAO Y G, SUI B K, ZHANG W.REL-SAGAN: Relative Generation Adversarial Network Integrated with Attention Mechanism for Scene Data Augmentation of Remote Sensing. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 2022, 15: 3107-3119. [10] LI S Y, SUNG Y.INCO-GAN: Variable-Length Music Generation Method Based on Inception Model-Based Conditional GAN. Mathematics, 2021, 9(4). DOI: 10.3390.math9040387. [11] INDRADI S D, ARIFIANTO A, RAMADHANI K N.Face Image Super-Resolution Using Inception Residual Network and GAN Framework // Proc of the 7th International Conference on Information and Communication Technology. Washington, USA: IEEE, 2019. DOI: 10.1109/ICoICT.2019.8835253. [12] BANA T, LOYA J, KULKARNI S.ViT-Inception-GAN for Image Colourisation // Proc of the 7th International Conference on Machine Learning, Optimization, and Data Science. Berlin, Germany: Springer, 2021: 105-118. [13] ZHENG J Y, WU Y H, SONG W R, et al. Multi-scale Feature Channel Attention Generative Adversarial Network for Face Sketch Synthesis. IEEE Access, 2021, 8: 146754-146769. [14] LI J W, LI B H, JIANG Y X, et al. MSAt-GAN: A Generative Adversarial Network Based on Multi-scale and Deep Attention Mechanism for Infrared and Visible Light Image Fusion. Complex and Intelligence Systems, 2022, 8: 4753-4781. [15] 汪璟玢,赖晓连,雷晶,等.基于注意力机制的多尺度空洞卷积神经网络模型.模式识别与人工智能, 2021, 34(6): 497-508. (WANG J B, LAI X L, LEI J, et al. Multi-scale Dilated Convolutional Neural Network Model Based on Attention Mechanism. Pattern Recognition and Artificial Intelligence, 2021, 34(6): 497-508.) [16] WOO S, PARK J, LEE J Y, et al. CBAM: Convolutional Block Attention Module // Proc of the 15th European Conference on Computer Vision. Berlin, Germany: Springer, 2018: 3-19. [17] SU B Y, ZHOU Z, CHEN H Y.PVEL-AD: A Large-Scale Open-World Dataset for Photovoltaic Cell Anomaly Detection. IEEE Transactions on Industrial Informatics, 2023, 19(1): 404-413. [18] WANG Q L, WU B G, ZHU P F, et al. ECA-Net: Efficient Channel Attention for Deep Convolutional Neural Networks // Proc of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. Washington, USA: IEEE, 2020: 11531-11539. [19] YANG L X, ZHANG R Y, LI L D, et al. SimAM: A Simple, Parameter-Free Attention Module for Convolutional Neural Networks // Proc of the 38th International Conference on Machine Learning. San Diego, USA: JMLR, 2021: 11863-11874. [20] HOU Q B, ZHOU D Q, FENG J S.Coordinate Attention for Efficient Mobile Network Design // Proc of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. Washington, USA: IEEE, 2021: 13708-13717. [21] ALEC R, LUKE M, SOUMITH C.Unsupervised Representation Learning with Deep Convolutional Generative Adversarial Networks[C/OL]. [2023-01-16].https://arxiv.org/pdf/1511.06434.pdf. [22] ARJOVSKY M, CHINTALA S, BOTTOU L, et al. Wasserstein Generative Adversarial Network // Proc of the 34th International Conference on Machine Learning. New York, USA: ACM, 2017: 214-223. [23] ZHANG H, GOODFELLOW I, METAXAS D, et al. Self-Attention Generative Adversarial Networks // Proc of the 36th International Conference on Machine Learning. San Diego, USA: JMLR, 2019: 7354-7363. [24] REN Y, DANG X W, WANG C, et al. Fast RCS Data Generation Based on infoGAN // Proc of Photonics and Electromagnetics Research Symposium. Washington, USA: IEEE, 2021: 331-336. [25] ARMANDPOUR M, SADEGHIAN A, LI C Y, et al. Partition-Guided GANs // Proc of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. Washington, USA: IEEE, 2021: 5095-5105. [26] KARRAS T, AITTALA M, LAINE S, et al. Alias-Free Generative Adversarial Networks // Proc of the 35th Conference on Neural Information Processing Systems[C/OL].[2023-01-16]. https://arxiv.org/pdf/2106.12423v1.pdf.
2023, Vol. 36 
