基于改进 YOLO v5算法的光伏组件红外热成像缺陷检测

孔松涛; 徐甄泽; 林星宇; 张椿秋; 蒋国庆; 张淳钦; 王堃

红外技术, 2023, 45 (9): 974, 网络出版: 2023-12-15

基于改进 YOLO v5算法的光伏组件红外热成像缺陷检测

Infrared Thermal Imaging Defect Detection of Photovoltaic Module Based on Improved YOLO v5 Algorithm

论文大纲

孔松涛 ^*徐甄泽林星宇张椿秋蒋国庆张淳钦王堃

作者单位

重庆科技学院机械与动力工程学院，重庆 401331

缺陷检测深度学习损失函数 SE模块平衡特征金字塔结构 defect detection, deep learning, YOLOv5, loss func YOLOv5

摘要

现有光伏组件缺陷识别方法存在提取特征困难、实时性较差导致了对光伏组件的缺陷故障检测的识别精度不高，本文提出一种基于改进 YOLO v5算法的光伏组件红外热成像缺陷检测方法。改进后的 YOLO v5算法主要是在原来的基础上增添注意机制 SE模块，并且改进损失函数将 GIoU改为 EIoU提高模型收敛效果、最后采用 KG模块平衡特征金字塔结构对模型进行优化，用以提高 YOLOv5算法的识别精度和收敛效果。改进后的网络结构应用在 YOLO v5s模型中，在光伏组件红外图像的检测上的平均检测精度 mAP可以达到 92.8%，比原本的 YOLO v5s算法 88.3%提升了 4.5%，在精确度和召回率上的收敛效果也比原始 YOLO v5算法模型有所提高，改进后的网络结构应用于 l、m、x三种模型中，其检测精度都有所提升，因此改进后的 YOLOv5算法适用于 4种模型。

Abstract

To solve the problem of difficulty in extracting features and poor real-time performance of existing photovoltaic power station defect identification methods, which lead to low identification accuracy of photovoltaic module defect detection, this paper proposes a photovoltaic power station infrared thermal imaging defect detection method based on an improved YOLO v5 algorithm. The improved YOLO v5 algorithm primarily adds an attention mechanism SE module to the original core and improves the loss function from GIoU to EIoU to enhance the model convergence effect. Finally, the knowledge graph (KG) module is used to balance the feature pyramid structure and optimize the model to improve the YOLO v5 algorithm’s recognition accuracy and convergence effects. The improved network structure was applied to the YOLO v5s model, whereby the average detection accuracy mAP used in the detection of infrared images of photovoltaic power plants reached 92.8%, which is 4.5% higher than that of the original YOLO v5s algorithm (88.3%). The effect of convergence on the precision and recall rate was also improved compared with the original YOLO v5 algorithm model. By applying the enhanced network structure to the three models (l, m, and x), detection accuracy was also improved. Consequently, the improved YOLO v5 algorithm is suitable for the four models.

参考文献

[1] 王丽霞 . 太阳能主动采暖室内热环境数值仿真 [J].计算机仿真 , 2021, 38(2): 5. WANG Lixia. Numerical simulation of indoor thermal environment of solar active heating[J]. Computer Simulation, 2021, 38(2): 5.

[2] 李伟. 太阳能光伏发电技术应用现状及未来发展趋势研究 [J]. 江苏科技信息, 2018, 35(24): 54-60. LI Wei. Research on application status and future development trend of solar photovoltaic power generation technology[J]. Jiangsu Science and Technology Information, 2018, 35(24): 54-60.

[3] GUO S, Schneller E, Davis K O, et al. Quantitative analysis of crystalline silicon wafer PV modules by electroluminescence imaging[C]//IEEE 43rd Photovoltaic Specialists Conference(PVSC), 2016: 3688-3692.

[4] Kandlikar S G, Perez Raya I, Raghupathi P A, et al. Infrared imaging technology for breast cancer detection–current status, protocols and new directions[J]. International Journal of Heat and Mass Transfer, 2017, 108: 2303-2320.

[5] ZHANG H, LUO C, WANG Q, et al. A novel infrared video surveillance system using deep learning based techniques[J]. Multimedia Tools and Applications, 2018, 77: 26657-26676.

[6] Cazzato D, Cimarelli C, Sanchez-Lopez J L, et al. A survey of computer vision methods for 2D object detection from unmanned aerial vehicles[J]. Journal of Imaging, 2020, 6(8): 78.

[7] Park J, CHEN J, Cho Y K, et al. CNN-based person detection using infrared images for night-time intrusion warning systems[J]. Sensors, 2019, 20(1): 34.

[8] Piniarski K, Paw.owski P. Efficient pedestrian detection with enhanced object segmentation in far IR night vision[C]//2017 Signal Processing: Algorithms, Architectures, Arrangements, and Applications (SPA) of IEEE, 2017: 160-165.

[9] LI S, WANG C, HUANG H. Infrared imaging guidance missile’s target recognition simulation based on air-to-air combat[C]//Optical Sensing and Imaging Technologies and Applications of SPIE, 2018, 10846: 768-780.

[10] JIAO L, ZHANG F, LIU F, et al. A survey of deep learning-based object detection[J]. IEEE Access, 2019, 7: 128837-128868.

[11] Krizhevsky A, Sutskever I, Hinton G E. ImageNet classification with deep convolutional neural networks[J]. Communications of the ACM, 2017, 60(6): 84-90.

[12] Thangaraj R, Anandamurugan S, Kaliappan V K. Automated tomato leaf disease classification using transfer learning-based deep convolution neural network[J]. Journal of Plant Diseases and Protection, 2021, 128(1): 73-86.

[13] Wenham P R, Price W H, Blundell G. Apolipoprotein E genotyping by one-stage PCR[J]. The Lancet, 1991, 337(8750): 1158-1159.

[14] Girshick R, Donahue J, Darrell T, et al. Rich feature hierarchies for accurate object detection and semantic segmentation[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 2017: 580-587.

[15] HE K, ZHANG X, REN S, et al. Spatial pyramid pooling in deep convolutional networks for visual recognition[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2015, 37(9): 1904-1916.

[16] Girshick R. Fast R-CNN[C]//Proceedings of the IEEE International Conference on Computer Vision, 2015: 1440-1448.

[17] REN S, HE K, Girshick R, et al. Faster R-CNN: towards real-time object detection with region proposal networks[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2015, 28: 1137-1149.

[18] Redmon J, Divvala S, Girshick R, et al. You only look once: unified, real-time object detection[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 2016: 779-788.

[19] Redmon J, Farhadi A. YOLO9000: better, faster, stronger[C]//

[20] Redmon J, Farhadi A. Yolov3: an incremental improvement[J/OL]. preprint arXiv:1804.02767. https://doi.org/10.48550/arXiv.1804.02767.

[21] Bochkovskiy A, WANG C Y, LIAO H Y M. Yolov4: Optimal speed and accuracy of object detection[J/OL]. [2020-04-23]. arXiv preprint arXiv: 2004.10934, https://doi.org/10.48550/arXiv.2004.10934.

[22] LIU W, Anguelov D, Erhan D, et al. SSD: Single shot multibox detector[C]// 14th European Conference of Computer Vision–ECCV, 2016: 21-37.

[23] Jocher G, Stoken Alex , Borovec J, et al. Ultralytics/yolov5: v5. 0-YOLOv5-P6 1280 models, AWS, Supervise. ly and YouTube integrations[Z/OL]. [2021-04-21] https://www.semanticscholar .org /paper/ultralytics-yolov5%3A-v5.0-YOLOv5-P6-1280-models%2C-and-Jocher-Stoken/ fd550b29c0efee 17be 5eb1447fddc 3c8ce66e838, Doi: 10.5281/ZENODO.4679653.

[24] 侯志强, 刘晓义, 余旺盛, 等. 使用 GIoU改进非极大值抑制的目标检测算法[J].电子学报 , 2021, 49(4): 10. HOU Zhiqiang, LIU Xiaoyi, YU Wangsheng et al. Improved object detection algorithm for non-maximum suppression using GIoU[J]. Acta Electronica Sinica, 2021, 49(4): 10.

[25] LI X, DING L, WANG L, et al. FPGA accelerates deep residual learning for image recognition[C]//2nd Information Technology, Networking, Electronic and Automation Control Conference(ITNEC) of IEEE, 2017: 837-840.

孔松涛, 徐甄泽, 林星宇, 张椿秋, 蒋国庆, 张淳钦, 王堃. 基于改进 YOLO v5算法的光伏组件红外热成像缺陷检测[J]. 红外技术, 2023, 45(9): 974. KONG Songtao, XU Zhenze, LIN Xingyu, ZHANG Chunqiu, JIANG Guoqing, ZHANG Chunqing, WANG Kun. Infrared Thermal Imaging Defect Detection of Photovoltaic Module Based on Improved YOLO v5 Algorithm[J]. Infrared Technology, 2023, 45(9): 974.

基于改进 YOLO v5算法的光伏组件红外热成像缺陷检测

关于本站 Cookie 的使用提示

全站搜索

基于改进 YOLO v5算法的光伏组件红外热成像缺陷检测

相关论文

相关资讯

关于本站 Cookie 的使用提示

全站搜索