A Robust Real-time Component for Personal Protective Equipment Detection in an Industrial Setting

In large industries, such as construction, metallurgy, and oil, workers are continually exposed to various hazards in their workplace. Accordingly to the International Labor Organization (ILO), there are 340 million occupational accidents annually. Personal Protective Equipment (PPE) is used to ensure the essential protection of workers’ health and safety. There is a great effort to ensure that these types of equipment are used properly. In such an environment, it is common to have closed-circuit television (CCTV) cameras to monitor workers, as those can be used to verify the PPE’s proper usage. Some works address this problem using CCTV images; however, they frequently can not deal with multiples safe equipment usage detection and others even skip the verification phase, making only the detection. In this paper, we propose a novel cognitive safety analysis component for a monitoring system. This component acts to detect the proper usage of PPE’s in real-time using data stream from regular CCTV cameras. We built the system component based on the top of state-of-art deep learning techniques for object detection. The methodology is robust with consistent and promising results for Mean Average Precision (80.19% mAP) and can act in real-time (80 FPS).

[1]  Jie Li,et al.  Safety helmet wearing detection based on image processing and machine learning , 2017, 2017 Ninth International Conference on Advanced Computational Intelligence (ICACI).

[2]  Amir H. Behzadan,et al.  Deep learning for site safety: Real-time detection of personal protective equipment , 2020 .

[3]  Geoffrey E. Hinton,et al.  ImageNet classification with deep convolutional neural networks , 2012, Commun. ACM.

[4]  Kaiming He,et al.  Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks , 2015, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[5]  Ali Farhadi,et al.  You Only Look Once: Unified, Real-Time Object Detection , 2015, 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[6]  Yang Bo,et al.  Helmet Detection under the Power Construction Scene Based on Image Analysis , 2019, 2019 IEEE 7th International Conference on Computer Science and Network Technology (ICCSNT).

[7]  Dietrich Paulus,et al.  Simple online and realtime tracking with a deep association metric , 2017, 2017 IEEE International Conference on Image Processing (ICIP).

[8]  Jixiu Wu,et al.  Automatic detection of hardhats worn by construction personnel: A deep learning approach and benchmark dataset , 2019, Automation in Construction.

[9]  Dumitru Erhan,et al.  Going deeper with convolutions , 2014, 2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[10]  Xianghua Zheng,et al.  Violation Monitoring System for Power Construction Site , 2019, IOP Conference Series: Earth and Environmental Science.

[11]  Jian Sun,et al.  Deep Residual Learning for Image Recognition , 2015, 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[12]  Xiaochun Luo,et al.  Detecting non-hardhat-use by a deep learning method from far-field surveillance videos , 2018 .

[13]  Andrew Zisserman,et al.  Very Deep Convolutional Networks for Large-Scale Image Recognition , 2014, ICLR.