Defect detection of light guide plate based on improved YOLOv5 networks

Abstract

Light guide plate (LGP) is a kind of material used in the backlight module. How to improve the quality control of LGP has become the focus of research in the industry. To address issues such as low gray contrast and a high proportion of small target defects in LGP images, an improved you only look once version 5 (YOLOv5) neural network based on multi-scale dilation convolution and a novel loss function is proposed. First, the LGP image is preprocessed, and then the context amplification module (CAM) is integrated into the feature fusion part of the detection algorithm to fuse multi-scale expansion convolution features to obtain rich context information. The extended intersection over union (XIoU) is selected as the location regression loss function. The results show that this method can effectively improve the detection accuracy and positioning accuracy. Compared with YOLOv5, the proposed method achieves an average accuracy increase of 4.7% and a recall rate increase of 2.7%. It can achieve accurate detection of defects, such as white/bright spots, black spots, line scratches, and surface foreign objects in LGP.

References

[1] MingW Y, ShenF, ZhangH M, et al.. Defect detection of LGP based on combined classifier with dynamic weights. Measurement, 2019, 143: 211-225 J]

[2] LiuF, LiJ F, DaiW Z. A defect detection method for light guide plates based on deep learning semantic segmentation. Computer system applications, 2020, 29(6):29-38[J]

[3] HuJ, LiJ F. Research on the scratch detection method of light guide plate line based on residual convolutional automatic encoder. Journal of optoelectronics·laser, 2020, 31(8):825-833[J]

[4] HuJ L, LiJ F. Defect detection of large-sized light guide plates based on improved YOLOv3. Computer system application, 2022, 31(06):279-286[J]

[5] LiuX, ZhangH Y. Defect detection of large-sized light guide plates based on improved YOLOv5s. Computer system application, 2023, 32(02):339-346[J]

[6] GLENN J. Ultralytics/YOLOv5 v6.0[EB/OL]. (2021-10-21) [2023-5-20]. https://github.com/ultralytics/yolov5/releases/tag/v6.0.

[7] XIAO J S, ZHAO T, YAO Y T, et al. Context augmentation and feature refinement network for tiny object detection[EB/OL]. (2022-04-14) [2023-5-30]. https://openreview.net/forum?id=q2ZaVU6bEsT.

[8] CSDN. New XIoU loss function improved in YOLOv7 and YOLOv5[EB/OL]. (2022-04-18) [2023-5-30]. https://yoloair.blog.csdn.net/article/details/130215696.

[9] LiJ F, LiM R. Research on defect detection methods for light guide plates based on machine vision. Journal of optoelectronics·laser, 2019, 30(3):256-265[J]

[10] HeY S. Research on quality visual inspection system of vehicle navigation light guide plate based on machine learning, 2021, Hangzhou, Zhejiang Sci-Tech University[D]

[11] ZhuX K, LuS C, WangX, et al.. TPH-YOLOv5: improved YOLOv5 based on transformer prediction head for object detection on drone-captured scenarios. Proceedings of the 2021 IEEE/CVF International Conference on Computer Vision, October 10–17, 2021, Montreal, QC, Canada, 2021, New York, IEEE: 2778-2788[C]

[12] TaoX, HouW, XuD. Review of surface defect detection methods based on deep learning. Journal of automation, 2021, 47(5):1017-1034[J]

[13] RenS Q, HeK M, GirshickR, et al.. Faster R-CNN: towards real-time object detection with region proposal networks. Proceedings of the 28th International Conference on Neural Information Processing Systems, December 7–12, 2015, Montreal, Canada, 2015, Cambridge, MIT Press: 91-99[C]

[14] LinT Y, GoyalP, GirshickR, et al.. Focal loss for dense object detection. 2017 IEEE International Conference on Computer Vision, October 24–27, 2017, Venice, Italy, 2017, New York, IEEE: 2999-3007[C]

[15] LiuW, AnguelovD, ErhanD, et al.. SSD: single shot multibox detector. Proceedings of the 14th European Conference on Computer Vision, October 11–14, 2016, Amsterdam, The Netherlands, 2016, Cham, Springer: 21-37[C]