YOLOv5 IS USED IN OPTIMIZATION OF SURFACE DEFECT DETECTION OF SOLAR CELLS

被引：0

作者：

Li, Yujuan ^{[1
,2
]}

Zhou, Jielong ^{[3
]}

Mai, Yaohua ^{[2
,4
]}

Wu, Shaohang ^{[2
,4
]}

Gao, Yanyan ^{[2
,4
]}

Li, Yang ^{[1
]}

机构：

[1] Faculty of Intelligent Manufacturing, Wuyi University, Jiangmen

[2] Guangdong Mellow Energy Co.,Ltd., Guangzhou

[3] Guangzhou Beihuan Intelligent Transportation Technology Co.,Ltd., Guangzhou

[4] Institute of New Energy Technology, Jinan University, Guangzhou

来源：

Taiyangneng Xuebao/Acta Energiae Solaris Sinica | 2024年 / 45卷 / 11期

关键词：

convolutional neural networks; deep learning; object detection; solar cells; YOLOv5;

D O I：

10.19912/j.0254-0096.tynxb.2023-1027

中图分类号：

学科分类号：

摘要：

In the field of new energy application technology,surface defect detection of solar cells is a crucial technical component. An optimized model based on the YOLOv5 algorithm is researched and proposed,which can detect static images and be used in real-time video. The model incorporates coordinate attention(CA)into the backbone network of YOLOv5 and optimizes the feature fusion part of the neck network using BiFPN structure. Furthermore,to address the issue of imbalanced positive and negative samples,Focal Loss is employed and optimized to Varifocal Loss. This optimized algorithm is applied to the PVEL-AD dataset,with mAP@0.5 used as the validation metric. The validation results demonstrate that the algorithm achieves a detection accuracy of 86.24%,which represents a 5.64% improvement compared to the unoptimized original algorithm. © 2024 Science Press. All rights reserved.

引用

页码：162 / 169

页数：7

共 25 条

[1]

GIRSHICK R, DONAHUE J, DARRELL T, Et al., Rich feature hierarchies for accurate object detection and semantic segmentation[C], 2014 IEEE Conference on Computer Vision and Pattern Recognition, pp. 580-587, (2014)

[2]

GIRSHICK R., Fast R-CNN, 2015 IEEE International Conference on Computer Vision(ICCV), pp. 1440-1448, (2015)

[3]

REN S Q, HE K M, 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, 39, 6, pp. 1137-1149, (2017)

[4]

HE K M, ZHANG X Y, REN S Q, Et al., Spatial pyramid pooling in deep convolutional networks for visual recognition[J], IEEE transactions on pattern analysis and machine intelligence, 37, 9, pp. 1904-1916, (2015)

[5]

LIU W, ANGUELOV D, ERHAN D, Et al., SSD:single shot MultiBox detector[M], Computer Vision – ECCV 2016, pp. 21-37, (2016)

[6]

REDMON J, DIVVALA S, GIRSHICK R, Et al., You only look once:unified,real-time object detection[C], 2016 IEEE Conference on Computer Vision and Pattern Recognition(CVPR), pp. 779-788, (2016)

[7]

REDMON J, FARHADI A., YOLO9000:better,faster, stronger, 2017 IEEE Conference on Computer Vision and Pattern Recognition(CVPR), pp. 6517-6525, (2017)

[8]

REDMON J, FARHADI A., YOLOv3: an incremental improvement [J], (2018)

[9]

BOCHKOVSKIY A, WANG C Y, LIAO H Y M., YOLOv4:optimal speed and accuracy of object detection [EB/OL], (2020)

[10]

LIN T Y, GOYAL P, GIRSHICK R, Et al., Focal loss for dense object detection [C], 2017 IEEE International Conference on Computer Vision (ICCV), pp. 2999-3007, (2017)

← 1 2 3 →