Abstract
To address the current problems of helmet and reflective vest detection algorithms, such as long inference times, high hardware requirements, and difficulty detecting small objects, an improved real-time detection algorithm based on YOLOv5 is proposed. First, the Mosaic-9 data enhancement method is adopted to expand the number of image splicings from four to nine, enriching the small-scale sample dataset. Minimum rectangular area cropping is used instead of the original random cropping to improve the training efficiency. Then, the EIOU (Efficient Intersection Over Union) loss is used as the loss function, and the focal loss is introduced to focus on the high-quality anchor box, achieving the purpose of accelerating the convergence speed and improving regression accuracy. Next, the CBAM-CA (Convolutional Block Attention Module—Coordinate Attention) hybrid attention mechanism is added to the network model. The purpose of the model is to focus on the small object’s spatial and channel feature information to improve the small object’s detection performance. Finally, a new scale object detection layer is added to learn the multilevel features of the dense object. The experimental results show that the improved model’s mAP (mean average precision) is 94.9%, an improvement of 7.0% over the original model. The size of the improved model is 4.49 MB, and the mean FPS (Frames Per Second) is 62. This algorithm guarantees a small model size while significantly increasing mAP, which is suitable for deployment on low computing power platforms.
Similar content being viewed by others
Availability of data and material
The datasets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request. Requests for material should be made to the corresponding authors.
References
Park, M.-W., Elsafty, N., Zhu, Z.: Hardhat-wearing detection for enhancing on-site safety of construction workers. J. Constr. Eng. Manage. 141(9), 04015024 (2015)
Yan, G.B., Sun, Q., Huang, J.Y., et al.: Helmet detection based on deep learning and random forest on UAV for power construction safety. J. Adv. Comput. Intell. Intell. Inf. 25(1), 40–49 (2021)
Luo, Q., Hua, J., Huang, L., et al.: Visualization analysis of domestic and overseas intelligent construction assisted by mapping knowledge domains. J. Build. Struct. 42(6), 1–14 (2021)
Liu, H., Zhang, F., Chen, Z., et al.: Applied research status and prospects of artificial intelligence in civil engineering field. J. Civil Environ. Eng. 66, 1–20 (2023). http://kns.cnki.net/kcms/detail/50.1218.TU.20220312.1605.002.html
Ramachandram, D., Taylor, G.W.: Deep multimodal learning: a survey on recent advances and trends. IEEE Signal Process. Mag. 34(6), 96–108 (2017)
Subramanyam, V., Kumar, J., Singh, S.N.: Temporal synchronization framework of machine-vision cameras for high-speed steel surface inspection systems. J. Real-Time Image Proc. 19(2), 445–461 (2022)
Le Cun, Y., Bengio, Y., Hinton, G.: Deep learning. Nature 521(7553), 436–444 (2015)
Wetzstein, G., Ozcan, A., Gigan, S., et al.: Inference in artificial intelligence with deep optics and photonics. Nature 588(7836), 39–47 (2020)
Zhou, Y., Zhang, L., Liu, T., et al.: Structural system identification based on computer vision. Chin. Civil Eng. J. 51(11), 17–23 (2018)
Zhu, X., Xiong, Y., Dai, J., et al.: Deep feature flow for video recognition. In: IEEE Conference on Computer Vision and Pattern Recognition. IEEE, pp. 4141–4150 (2017)
Yang, S.D., Chen, Z.H., Ma, X.M., et al.: Real-time high-precision pedestrian tracking: a detection-tracking-correction strategy based on improved SSD and Cascade R-CNN. J. Real-Time Image Proc. 19(2), 287–302 (2021)
Ren, S., He, K., Girshick, R., et al.: Faster R-CNN: towards real time object detection with region proposal networks. IEEE Trans. Pattern Anal. Mach. Intell. 39(6), 1137–1149 (2017)
Pan, H.G., Shi, Y.H., Lei, X.Y., et al.: Fast identification model for coal and gangue based on the improved tiny YOLO v3. J. Real-Time Image Proc. 19(3), 687–701 (2022)
Jain, V., Jadhav, N., Verhelst, M.: Enabling real-time object detection on low cost FPGAs. J. Real-Time Image Proc. 19(1), 217–229 (2021)
Lin, J., Dang, W.C., Pan, L.H., et al.: Safety helmet detection based on YOLO. Comp. Syst. Appl. 28(9), 174–179 (2019)
Redmon, J., Divvala, S., Girshick, R., et al.: You only look once: unified, real-time object detection. In: IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Seattle, WA, pp. 779–788 (2016)
Redmon, J., Farhadi, A.: YOLO9000: better, faster, stronger. In: IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Honolulu, HI, pp. 6517–6525 (2017)
Fang, M., Sun, T.T., Shao, Z.: Fast helmet-wearing-condition detection based on improved YOLOv2. Opt. Precis. Eng. 27(5), 1196–1205 (2019)
Yang, L.Q., Cai, L.Q., Gu, S.: Detection on wearing behavior of safety helmet based on machine learning method. J. Saf. Sci. Technol. 15(10), 152–157 (2019)
Shi, H., Chen, X., Yang, Y.: Safety helmet wearing detection method of improved YOLO v3. Comput. Eng. Appl. 55(11), 213–220 (2019)
Wu, F., Jin, G.Q., Gao, M.Y., et al.: Helmet detection based on improved YOLOV3 deep model. In: 2019 IEEE 16th International Conference on Networking, Sensing and Control (ICNSC), Banff, AB, Canada, pp 363–368 (2019)
Wang, Y., Gu, Y., Feng, X., et al.: Research on detection method of helmet wearing based on attitude estimation. Appl. Res. Comput. 38(3), 937–940 (2021)
Xu, C., Yuan, H., Li, G., et al.: Research on safety helmet wearing detection method based on scene augment. Comput. Eng. Appl. 58(19), 326–332 (2022)
Zhu, X., Lyu, S., Wang, X., et al.: TPH-YOLOv5: improved YOLOv5 based on transformer prediction head for object detection on drone-captured scenarios. In: IEEE/CVF International Conference on Computer Vision (ICCVW), Electr Network, pp. 2778–2788 (2021)
Zhu, L.L., Geng, X., Li, Z., et al.: Improving YOLOv5 with attention mechanism for detecting boulders from planetary images. Remote Sens. 13(18), 152–161 (2021)
Luo, S., Yu, J., Xi, Y.J., et al.: Aircraft target detection in remote sensing images based on improved YOLOv5. IEEE Access 10, 5184–5192 (2022)
Chen, J.Z., Jia, K.K., Chen, W.Q., et al.: A real-time and high-precision method for small traffic-signs recognition. Neural Comput. Appl. 34(3), 2233–2245 (2021)
Fang, W., Wang, L., Ren, P.M.: Tinier-YOLO: a real-time object detection method for constrained environments. IEEE Access 8, 1935–1944 (2019)
Zhang, G.Q., Yang, J.C., Zheng, Y.H., et al.: Hybrid-attention guided network with multiple resolution features for person reidentification. Inf. Sci. 578(8), 51–62 (2021)
Woo, S., Park, J., Lee, J.-Y., et al.: CBAM: convolutional block attention module. In: 15th European Conference on Computer Vision (ECCV), Munich, Germany, pp. 3–19 (2018)
Hou, Q.B., Zhou, D.Q., Feng, J.S.: Coordinate attention for efficient mobile network design. In: IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), Electr Network, pp. 13708–13717 (2021)
Acknowledgements
This research is supported by Hebei Province Full-time Top-level Talents Introduction Project, No. 2020HBQZYC013 and the China Construction Sixth Engineering Bureau Research and Development Project, No. CSCEC6B-2022-Z-2. We are grateful to the State Key Laboratory of Hydraulic Engineering Simulation and Safety of Tianjin University for its support to this project. Thanks to Elsevier language editing services.
Funding
Hebei Province Full-time Top-level Talents Introduction Project, 2020HBQZYC013, Hongbo Liu, China Construction Sixth Engineering Bureau Research and Development Project, CSCEC6B-2022-Z-2, Qian Zhang.
Author information
Authors and Affiliations
Corresponding author
Ethics declarations
Conflict of interest
The authors declare that they have no competing interests.
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
About this article
Cite this article
Chen, Z., Zhang, F., Liu, H. et al. Real-time detection algorithm of helmet and reflective vest based on improved YOLOv5. J Real-Time Image Proc 20, 4 (2023). https://doi.org/10.1007/s11554-023-01268-w
Received:
Accepted:
Published:
DOI: https://doi.org/10.1007/s11554-023-01268-w