International Journal of Innovative Research in Computer and Communication Engineering
ISSN Approved Journal | Impact factor: 8.771 | ESTD: 2013 | Follows UGC CARE Journal Norms and Guidelines
| Monthly, Peer-Reviewed, Refereed, Scholarly, Multidisciplinary and Open Access Journal | High Impact Factor 8.771 (Calculated by Google Scholar and Semantic Scholar | AI-Powered Research Tool | Indexing in all Major Database & Metadata, Citation Generator | Digital Object Identifier (DOI) |
| TITLE | Monitoring Safety and Compliance in the Industrial Sector using Computer Vision and Deep Learning Model |
|---|---|
| ABSTRACT | Ensuring worker safety on factory floors, construction sites, and process plants remains a persistent challenge, with the International Labour Organization reporting over 2.3 million work-related deaths annually and non-compliance with Personal Protective Equipment (PPE) policy identified as a leading contributing factor. Traditional safety monitoring relies on manual supervision and periodic audits, which are labour-intensive, inconsistent, and unable to provide continuous coverage across large industrial premises. This paper presents a real-time Industrial Safety and Compliance Monitoring System that applies computer vision and deep learning to automatically detect PPE usage, identify safety violations, and flag unauthorized entry into restricted zones directly from CCTV video feeds. The system is built around a custom-trained YOLOv8-based object detection model that identifies workers and classifies seven safety-relevant classes — Helmet, No-Helmet, Safety-Vest, No-Vest, Gloves, Restricted-Zone-Person, and Fire-Extinguisher-Access-Blocked. Trained on an aggregated dataset of 12,600 annotated industrial images, the proposed model achieves a mean Average Precision ([email protected]) of 93.8%, Precision of 92.4%, Recall of 90.6%, and a real-time inference speed of 41 FPS on a single GPU. Grad-CAM-based visual explainability is integrated to highlight the exact image regions driving each detection, and a rule-based compliance-scoring engine converts frame-level detections into a live safety score per zone. The complete system is deployed as a Flask web dashboard with real-time video overlay, automated violation alerts, and historical compliance analytics, enabling safety officers to intervene before incidents occur rather than after they are reported. |
| AUTHOR | K. SHANKAR, SAI KRISHNA DALAI, L. LALITHA Sr. Asst. Professor, Dept. of CSE, NSRIT, Visakhapatnam, AP, India Asst. Professor, Dept. of MCA, NSRIT, Visakhapatnam, AP, India Dept. of MCA, NSRIT, Visakhapatnam, AP, India |
| VOLUME | 186 |
| DOI | DOI: 10.15680/IJIRCCE.2026.1407059 |
| pdf/59_Monitoring Safety and Compliance in the Industrial Sector using Computer Vision and Deep Learning Model.pdf | |
| KEYWORDS | |
| References | [1] D. Nath, A. H. Behzadan, and S. G. Paal, "Deep learning for site safety: Real-time detection of personal protective equipment," Automation in Construction, vol. 112, 2020. [2] Z. Wang, Y. Wu, L. Yang, A. Thirunavukarasu, C. Evison, and Y. Zhao, "Fast personal protective equipment detection for real construction sites using deep learning approaches," Sensors, vol. 21, no. 10, 2021. [3] S. S. Delhi, R. Sankarlal, and A. Thomas, "Detection of personal protective equipment (PPE) compliance on construction site using computer vision based deep learning techniques," Frontiers in Built Environment, vol. 6, 2020. [4] R. Xiong and P. Tang, "Machine learning using synthetic images for detecting and locating hardhats worn by construction workers," Journal of Computing in Civil Engineering, vol. 35, no. 1, 2021. [5] K. Selvaraj and V. Adithya, "A survey on vision-based occupational safety and health monitoring systems," International Journal of Industrial Ergonomics, vol. 89, 2022. [6] A. Selvam and R. Rangarajan, "Explainable deep learning for industrial visual inspection using Grad-CAM," Procedia CIRP, vol. 107, 2022. [7] J. Redmon, S. Divvala, R. Girshick, and A. Farhadi, "You Only Look Once: Unified, real-time object detection," in Proc. IEEE CVPR, Las Vegas, NV, 2016, pp. 779–788. [8] G. Jocher et al., "Ultralytics YOLOv8," GitHub repository, 2023. Available: https://github.com/ultralytics/ultralytics [9] W. Liu et al., "SSD: Single shot multibox detector," in Proc. ECCV, Amsterdam, 2016, pp. 21–37. [10] S. Ren, K. He, R. Girshick, and J. Sun, "Faster R-CNN: Towards real-time object detection with region proposal networks," IEEE TPAMI, vol. 39, no. 6, pp. 1137–1149, 2017. [11] R. R. Selvaraju, M. Cogswell, A. Das, R. Vedantam, D. Parikh, and D. Batra, "Grad-CAM: Visual explanations from deep networks via gradient-based localization," in Proc. IEEE ICCV, Venice, 2017, pp. 618–626. [12] International Labour Organization, "Safety and health at work: A vision for sustainable prevention," ILO Report, Geneva, 2023. [13] Occupational Safety and Health Administration (OSHA), "Personal Protective Equipment Standards," 29 CFR 1910 Subpart I, U.S. Department of Labor. [14] International Organization for Standardization, "ISO 45001:2018 — Occupational health and safety management systems." [15] Roboflow, "Hard Hat Workers Dataset," Available: https://public.roboflow.com/object-detection/hard-hat-workers [16] SHEL5K Dataset, "Safety Helmet and Vest Detection Dataset," Available: https://github.com/ZijianWang1995/SHEL5K [17] Ultralytics Documentation. Available: https://docs.ultralytics.com [18] OpenCV Documentation. Available: https://docs.opencv.org [19] Flask Documentation. Available: https://flask.palletsprojects.com [20] PyTorch Documentation. Available: https://pytorch.org/docs [21] K. He, X. Zhang, S. Ren, and J. Sun, "Deep residual learning for image recognition," in Proc. IEEE CVPR, Las Vegas, NV, 2016, pp. 770–778. [22] A. Bochkovskiy, C.-Y. Wang, and H.-Y. M. Liao, "YOLOv4: Optimal speed and accuracy of object detection," arXiv preprint arXiv:2004.10934, 2020. [23] T.-Y. Lin et al., "Feature pyramid networks for object detection," in Proc. IEEE CVPR, Honolulu, HI, 2017, pp. 2117–2125. [24] T.-Y. Lin et al., "Microsoft COCO: Common objects in context," in Proc. ECCV, Zurich, 2014, pp. 740–755. [25] I. Goodfellow, Y. Bengio, and A. Courville, Deep Learning, MIT Press, 2016. [26] R. Girshick, "Fast R-CNN," in Proc. IEEE ICCV, Santiago, 2015, pp. 1440–1448. [27] D. P. Kingma and J. Ba, "Adam: A Method for Stochastic Optimization," in Proc. ICLR, San Diego, CA, 2015. [28] S. Ioffe and C. Szegedy, "Batch Normalization: Accelerating deep network training by reducing internal covariate shift," in Proc. ICML, Lille, 2015, pp. 448–456. [29] C.-Y. Wang, A. Bochkovskiy, and H.-Y. M. Liao, "YOLOv7: Trainable bag-of-freebies sets new state-of-the-art for real-time object detectors," in Proc. IEEE CVPR, 2023, pp. 7464–7475. [30] Z. Zheng et al., "Distance-IoU loss: Faster and better learning for bounding box regression," in Proc. AAAI, 2020, pp. 12993–13000. [31] IEEE Xplore — Computer Vision for Occupational Safety. Available: https://ieeexplore.ieee.org [32] Google Scholar — PPE Detection Deep Learning Research. Available: https://scholar.google.com [33] National Safety Council, "Work Injury Costs and Statistics," Injury Facts Report, 2023. [34] F. Chollet, Deep Learning with Python, 2nd ed., Manning Publications, 2021. [35] A. Géron, Hands-On Machine Learning with Scikit-Learn, Keras, and TensorFlow, 3rd ed., O'Reilly Media, 2022. [36] S. Haykin, Neural Networks and Learning Machines, 3rd ed., Pearson, 2009. [37] N. Srivastava, G. Hinton, A. Krizhevsky, I. Sutskever, and R. Salakhutdinov, "Dropout: A simple way to prevent neural networks from overfitting," JMLR, vol. 15, pp. 1929–1958, 2014. [38] M. Everingham et al., "The PASCAL Visual Object Classes (VOC) challenge," IJCV, vol. 88, pp. 303–338, 2010. [39] Python Documentation. Available: https://docs.python.org [40] NVIDIA Jetson Developer Documentation. Available: https://developer.nvidia.com/embedded-computing |
