Detection of fires at technological facilities using convolutional neural network

Introduction. The article considers the solution of the problem of detecting fires at technological facilities in automatic mode. To solve this problem, an approach was chosen to create a convolutional neural network capable of operating on a real-time video stream.

Aims and Purposes. The aim of this work is to create a neural network capable of detecting flames and smoke in the image from CCTV cameras. The purposes of the work include: selection of the optimal architecture of the neural network in accordance with the latest research in this field; speeding up the chosen architecture using quantization and filter thinning techniques.

Methods. Different architectures of convolutional neural networks performing the task of detecting objects in an image are considered. Their performance and quality of work are compared. The YOLOv5 architecture, its target function, training methods and ways of speeding up work are considered.

Results and discussion. The paper shows the results of training a convolutional neural network of the YOLOv5 architecture for the task of flame and smoke detection, as well as how the results change when applying neural network acceleration methods. It was determined that the use of such acceleration methods such as quantization and filter cleaning can significantly increase up the speed of the neural network, while almost no loss in accuracy of operation.

Conclusions. As a result of the conducted work, the architecture of the neural network was determined for performing the task of detecting a fire source. Based on the chosen architecture, a neural network was trained to detect flames and smoke in the image. The speed of its work allows to process video stream in real-time without using graphic accelerator.

1. Mazur-Milecka M., Glowacka N., Kaczmarek M., Bujnowski A., Kaszynski M., RuminskiSmart J. City and fire detection using thermal imaging. 14th International Conference on Human System Interaction (HSI). 2021; 1-7. DOI: 10.1109/HSI52170.2021.9538699

2. Chacon M., Perez-Vargas F. Thermal Video Analysis for Fire Detection Using Shape Regularity and Intensity Saturation Features. Third Mexican Conference MCPR. 2011; 118-126. DOI: 10.1007/978-3-642-21587-2_13

3. Zaman T., Hasan M., Ahmed S., Ashfaq S. Fire Detection Using Computer Vision. IEEE 61st International Midwest Symposium on Circuits and Systems (MWSCAS). 2018; 356-359. DOI: 10.1109/MWSCAS.2018.8623842

4. Manjunatha K., Mohana H., Vijaya P. Implementation of Computer Vision Based Industrial Fire Safety Automation by Using Neuro-Fuzzy Algorithms. I.J. Information Technology and Computer Science. 2015; 14-27. DOI: 10.5815/ijitcs.2015.04.02

5. Mondal M., Prasad V., Kumar R., Saha N., Saumadeep G., Ratna G., Mukhopadhyay A., Sourav S. Automating Fire Detection and Suppression with Computer Vision: A Multi-Layered Filtering Approach to Enhanced Fire Safety and Rapid Response. Fire Technol. 2023. DOI: 10.1007/s10694-023-01392-w

6. Girshick R., Donahue J., Darrell T., Malik J. Rich Feature Hierarchies for Accurate Object Detection and Semantic Segmentation. 2014 IEEE Conference on Computer Vision and Pattern Recognition. 2014; 580-587. DOI: 10.1109/CVPR.2014.81

7. He K., Gkioxari G., Dollar P., Girshick R. Mask R-CNN. 2017 IEEE International Conference on Computer Vision (ICCV). 2017; 2980-2988. DOI: 10.1109/ICCV.2017.322

8. Lin T., Dollar P., Girshick R., He K., Hariharan B., Belongie S. Feature Pyramid Networks for Object Detection. 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR). 2017; 936-944. DOI: 10.1109/CVPR.2017.106

9. Liu Z. et al. Swin Transformer: Hierarchical Vision Transformer using Shifted Windows. 2021 IEEE/CVF International Conference on Computer Vision (ICCV). 2021; 9992-10002. DOI: 10.1109/ICCV48922.2021.00986

10. Redmon J., Divvala S., Girshick R., Farhadi A. You Only Look Once: Unified, Real-Time Object Detection. CVPR. 2016; 779-788. DOI: 10.1109/CVPR.2016.91

11. Liu W. et al. SSD: Single Shot MultiBox Detector. Computer Vision — ECCV Lecture Notes in Computer Science. 2016; 9905. DOI: 10.1007/978-3-319-46448-0_2

12. Lin T., Goyal P., Girshick R., He K., Dollar P. Focal Loss for Dense Object Detection. 2020 IEEE Transactions on Pattern Analysis and Machine Intelligence. 2020; 42(2):318-327. DOI: 10.1109/TPAMI.­­2018.2858826

13. Lin T. et al. Microsoft COCO: Common Objects in Context. ECCV 2014. Lecture Notes in Computer Science. 2014; 8693. DOI: 10.1007/978-3-319-10602-1_48

14. Sobti A., Arora C., Balakrishnan M. Object Detection in Real-Time Systems: Going Beyond Precision. 2018 IEEE Winter Conference on Applications of Computer Vision (WACV). 2018; 1020-1028. DOI: 10.1109/WACV.2018.00117

15. Henderson P., Ferrari V. End-to-End Training of Object Class Detectors for Mean Average Precision. ACCV 2016. Lecture Notes in Computer Science. 2016; 10115. DOI: 10.1007/978-3-319-54193-8_13

16. Li C., Li L., Geng Y., Jiang H., Cheng M., Zhang B., Ke Z., Xu X., Chu X. YOLOv6 v3.0: A Full-Scale Reloading. arXiv: Computer Science. 2023. DOI: 10.48550/arXiv.2301.05586

17. Wang C., Yeh I., Liao H. You Only Learn One Representation: Unified Network for Multiple Tasks. Journal of Information Science and Engineering. 2021; 39(2):691-709. DOI: 10.48550/arXiv.2105.04206

18. Wang C., Bochkovskiy A., Liao H. Scaled-YOLOv4: Scaling Cross Stage Partial Network. 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). 2021; 13024-13033. DOI: 10.1109/CVPR46437.2021.01283

19. Redmon J., Farhadi A. YOLOv3: An Incremental Improvement. arXiv: Computer Science. 2018. DOI: 10.48550/arXiv.1804.02767

20. Ramachandran P., Zoph B., Le Q. Swish: a Self-Gated Activation Function. arXiv: Neural and Evolutionary Computing. 2017. DOI: 10.48550/arXiv.1710.05941

21. Robbins H. A Stochastic Approximation Method. Annals of Mathematical Statistics. 1951; 22:400-407. DOI: 10.1214/aoms/1177729586

22. Kingma D., Ba J. Adam: A Method for Stochastic Optimization. arXiv: Computer Science. 2014. DOI: 10.48550/arXiv.1412.6980

23. Zhang Z. Improved Adam Optimizer for Deep Neural Networks. 2018 IEEE/ACM 26th International Symposium on Quality of Service (IWQoS). 2018; 1-2. DOI: 10.1109/IWQoS.2018.8624183