Detection Method of Safety Protective Equipment Based on Deep Learning

Authors

  • Zhu Shi
  • Hao Wu
  • Zhongyang Jin
  • Hong Song

DOI:

https://doi.org/10.54097/ajst.v4i2.3897

Keywords:

Safety appliances, YOLOv5, BoT, Self attention mechanism.

Abstract

The construction site belongs to the high-risk operation area, and whether the safety protection articles are correctly worn will directly affect the personal safety of operators. Aiming at the problems of weak anti-interference and low detection accuracy of existing detection methods, a new detection method for safety appliances based on improved YOLOv5 is proposed. This method embeds the BoT module into YOLOv5 backbone network, and uses its self attention mechanism to further mine image feature clues, so as to enhance the network's ability to extract features from target regions. The experimental results show that the average precision mAP of the method based on the improved YOLOv5 is 87.1%, and the recall rate is 80.8%. Compared with the original YOLOv5, the precision is improved by two percent.

Downloads

Download data is not yet available.
<br data-mce-bogus="1"> <br data-mce-bogus="1">

References

Ge Qingqing, Zhang Zhijie, Yuan Long, Li Xiumei, Sun Junmei. Safety helmet wearing detection by integrating environmental features and improving YOLOv4 [J]. Chinese Journal of Image Graphics, 2021,26 (12): 2904-2917.

Liu Yunbo, Huang Hua. Research on Monitoring Technology of Helmet Wearing on Construction Site [J]. Electronic Science and Technology, 2015, 28 (04): 69-72.

Rubaiyat A H M, Toma T T, Kalantari-Khandani M, et al. Automatic detection of helmet uses for construction safety[C]//2016 IEEE/WIC/ACM International Conference on Web Intelligence Workshops (WIW). IEEE, 2016: 135-142.

Wang Zhen, Jin Yong, Wang Zhaoba, Chen Youxing. Workwear detection and tracking based on dual weight color histogram [J]. TV Technology, 2016, 40 (01): 131-134.

Zhong Xinhao Research on Real time Monitoring Method of Helmet Wearing in Metro Construction Scenarios [D]. Hunan University of Technology, 2021.

Li Mingshan, Han Qingpeng, Zhang Tianyu, Wang Daolei. Improving the safety helmet detection method of SSD [J]. Computer Engineering and Application, 2021,57 (08): 192-197.

Zhang Wukang, Pan Lizhi, Guo Zhibin, Lin Xuan, Tu Xiaotong. RetinaNet based visual detection method for abnormal state of insulating gloves in power scenarios [J]. Journal of Hunan University of Science and Technology (Natural Science Edition), 2022,37 (01): 85-91.

Zheng Haiyang, Song Chunhe, Wu Tingting, Liu Shuo, Zhou Zhongran. Small target detection and matching algorithm for insulating gloves wear detection [J/OL]. Small microcomputer system: 1-10 [2022-10-11].

Srinivas A, Lin T Y, Parmar N, et al. Bottleneck transformers for visual recognition[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. 2021: 16519-16529.

Dosovitskiy A, Beyer L, Kolesnikov A, et al. An image is worth 16x16 words: Transformers for image recognition at scale[J]. arXiv preprint arXiv:2010.11929, 2020.

VASWANI A, SHAZEER N, PARMAR N, et al. Attention is all you need[C]//Proceedings of the 31st International Conference on Neural Information Processing Systems Long Beach, CA, Dec 4-9, 2017.Cambridge: The MIT Press, 2017: 6000-6010. [5]SHI Biao, LI Yu Xia, YU Xhua, YAN Wang. Short-term load forecasting based on modified particle swarm optimizer and fuzzy neural network model [J]. Systems Engineering-Theory and Practice, 2010, 30(1): 158-160.

Downloads

Published

04-01-2023

How to Cite

Shi, Z., Wu, H., Jin, Z., & Song, H. (2023). Detection Method of Safety Protective Equipment Based on Deep Learning. Academic Journal of Science and Technology, 4(2), 45–49. https://doi.org/10.54097/ajst.v4i2.3897

Issue

Section

Articles