Intelligent Monitoring of Tunnel Fire Smoke Based on Improved YOLOX and Edge Computing
Source abstract
Abstract
To overcome the defects of traditional fire detection methods that have a high false alarm rate and long delay, a smart tunnel fire monitoring method based on a YOLOX deep convolutional neural network and edge computing is proposed. This method first improves the detection accuracy by analyzing the relationship between frequency domain and convolutional neural networks and the use of wavelet transform. Then, based on the smoke features observed in the experiments, a fuzzy loss method is proposed to accelerate the model convergence speed. To address the issue of a weak computing power of edge devices, the training model is optimized by using knowledge distillation and model quantization, thereby improving the running speed on edge devices. At the same time, a series of related lightweight methods are adopted to optimize the model, reduce the computational cost, and improve the detection speed. Finally, the accuracy of flame and smoke detection on a self-built dataset reaches 85%, which is about 1.8% higher than the baseline method YOLOX and achieves a balance between the speed and accuracy of the model.
Source: crossref · Quality: baseline
Research context
OpenAlex topics: Fire Detection and Safety Systems
SciVX discovery tags are terminology-based. OpenAlex topics are source-provided classifications.
Affiliations
Departement of Civil Engineering, Southeast University, Nanjing 211189, China
College of Software Engineering, Southeast University, Suzhou 215000, China
Funding
National Natural Science Foundation of China
References (59)
Show source-provided references
1. Ministry of Transport of China (2023, June 25). 2020 China National Toll Road Statistical Bulletin, Available online: https://xxgk.mot.gov.cn/2020/jigou/glj/202110/t20211027_3623195.html.
2. Safety evaluation of lighting at very long tunnels on the basis of visual adaptation10.1016/j.ssci.2019.03.018 ↗
3. An empirical investigation of driver’s eye-catching effect in the entrance zone of freeway tunnels: A naturalistic driving experiment10.1016/j.trf.2024.01.004 ↗
4. Tunnel safety: A pilot study investigating drivers’ fixation characteristics when approaching tunnel entrance at different driving speeds10.1177/14771535221100261 ↗
5. Risk analysis in road tunnels10.1016/j.proeng.2014.12.028 ↗
6. Overview of traffic safety aspects and design in road tunnels10.1016/j.iatssr.2016.02.002 ↗
7. Road tunnel fire safety and risk: A review10.1186/s40038-015-0006-6 ↗
8. Fire evacuation strategy analysis in long metro tunnels10.1016/j.ssci.2021.105603 ↗
9. Fire safety in tunnels10.1016/j.firesaf.2008.06.008 ↗
10. CFD analysis of the dispersion of toxic materials in road tunnels10.1016/j.jlp.2013.05.002 ↗
11. Smart real-time forecast of transient tunnel fires by a dual-agent deep learning model10.1016/j.tust.2022.104631 ↗
12. A multi-neural network fusion algorithm for fire warning in tunnels10.1016/j.asoc.2022.109799 ↗
13. Fire detection based on vision sensor and support vector machines10.1016/j.firesaf.2008.07.006 ↗
14. Xiang, X.B. (2016). The Research of Smoke Detection Algorithm on Video, Zhejiang University. (In Chinese).
15. Yang, C. (2020). Research and Development of Smoke Video Detection System Based on Deep Learning, Nanjing University of Posts and Telecommunications. (In Chinese).
16. Lei, D. (2017). Research on Highway Tunnel Smoke Detection Method Based on Video, Dalian Maritime University. (In Chinese).
17. Li, Y.Y. (2019). Study on Flame and Smoke Detection Method Based on Significance in Early Stage of Fire, Dalian Maritime University. (In Chinese).
18. Video Smoke Detection Based on YUV Color Space and Multiple Feature Fusion
19. Adaptive smoke detection algorithm based on multi-feature fusion
20. Early smoke detection of forest fires based on SVM image segmentation10.17221/82/2018-JFS ↗
21. An effective algorithm to detect both smoke and flame using color and wavelet analysis10.1134/S1054661817010138 ↗
22. Edge computing: Vision and challenges10.1109/JIOT.2016.2579198 ↗
23. The emergence of edge computing10.1109/MC.2017.9 ↗
24. Deep learning with edge computing: A review10.1109/JPROC.2019.2921977 ↗
25. Li, M., Bi, Z., Wang, T., Wen, Y., Niu, Q., Liu, J., Peng, B., Jang, S., Pan, X., and Xu, J. (2024). Deep learning and machine learning with gpgpu and cuda: Unlocking the power of parallel computing. arXiv.
26. Cheng, Y., Wang, D., Zhou, P., and Zhang, T. (2017). A survey of model compression and acceleration for deep neural networks. arXiv.
27. Early fire detection using convolutional neural networks during surveillance for effective disaster management10.1016/j.neucom.2017.04.083 ↗
28. Early smoke and flame detection based on transformer
29. Toreyin, B.U., Dedeoglu, Y., and Cetin, A.E. (2006, January 4–8). Contour based smoke detection in video using wavelets. Proceedings of the 2006 14th European Signal Processing Conference, Florence, Italy.
30. Fire detection in video sequences using a generic color model10.1016/j.firesaf.2008.05.005 ↗
31. Shafi, I., Mazhar, M.F., Fatima, A., Alvarez, R.M., Miró, Y., Espinosa, J.C.M., and Ashraf, I. (2023). Deep learning-based real time defect detection for optimization of aircraft manufacturing and control performance. Drones, 7.10.3390/drones7010031 ↗
32. Modified densenet for automatic fabric defect detection with edge computing for minimizing latency10.1109/JIOT.2020.2983050 ↗
33. Wang, H., and Vieira, J. (2010, January 7–9). 2-D wavelet transforms in the form of matrices and application in compressed sensing. Proceedings of the 2010 8th World Congress on Intelligent Control and Automation, Jinan, China.
34. Lin, T., Goyal, P., Girshick, R.B., He, K., and Dollar, P. (2017, January 22–29). Focal loss for dense object detection. Proceedings of the IEEE International Conference on Computer Vision, Venice, Italy.10.1109/ICCV.2017.324 ↗
35. Yu, J., Jiang, Y., Wang, Z., Cao, Z., and Huang, T. (2016, January 15–19). Unitbox: An advanced object detection network. Proceedings of the 24th ACM International Conference on Multimedia, Amsterdam, The Netherlands.10.1145/2964284.2967274 ↗
36. Rezatofighi, H., Tsoi, N., Gwak, J.Y., Sadeghian, A., Reid, I., and Savarese, S. (2019, January 15–20). Generalized intersection over union: A metric and a loss for bounding box regression. Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, Long Beach, CA, USA.10.1109/CVPR.2019.00075 ↗
37. Zheng, Z., Wang, P., Liu, W., Li, J., Ye, R., and Ren, D. (2020, January 7–12). Distance-IoU loss: Faster and better learning for bounding box regression. Proceedings of the AAAI Conference on Artificial Intelligence, New York, NY, USA.10.1609/aaai.v34i07.6999 ↗
38. Focal and efficient IOU loss for accurate bounding box regression10.1016/j.neucom.2022.07.042 ↗
39. Gonzales, R.C., and Wintz, P. (1987). Digital Image Processing, Addison-Wesley Longman Publishing Co., Inc.
40. (2014). Detailed Regulations for Highway Tunnel Lighting Design (Standard No. JTG/T D70/2-01-2014).
41. Ge, Z., Liu, S., Wang, F., Li, Z., and Sun, J. (2021). Yolox: Exceeding yolo series in 2021. arXiv.
42. Howard, A.G., Zhu, M., Chen, B., Kalenichenko, D., Wang, W., Weyand, T., Adreetto, M., and Adam, H. (2017). Mobilenets: Efficient convolutional neural networks for mobile vision applications. arXiv.
43. Spatial pyramid pooling in deep convolutional networks for visual recognition10.1109/TPAMI.2015.2389824 ↗
44. MMDetection Contributors (2023, June 26). OpenMMLab Detection Toolbox and Benchmark. Available online: https://github.com/open-mmlab/mmdetection.
45. Hinton, G. (2015). Distilling the Knowledge in a Neural Network. arXiv.
46. Deep learning10.1038/nature14539 ↗
47. He, K., Zhang, X., Ren, S., and Sun, J. (2016, January 27–30). Deep residual learning for image recognition. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, Las Vegas, NV, USA.10.1109/CVPR.2016.90 ↗
48. Language models are few-shot learners
49. Han, S., Mao, H., and Dally, W.J. (2015). Deep compression: Compressing deep neural networks with pruning, trained quantization and huffman coding. arXiv.
50. Knowledge distillation: A survey10.1007/s11263-021-01453-z ↗
51. Romero, A., Ballas, N., Kahou, S.E., Chassang, A., Gatta, C., and Bengio, Y. (2014). Fitnets: Hints for thin deep nets. arXiv.
52. Zagoruyko, S., and Komodakis, N. (2016). Paying more attention to attention: Improving the performance of convolutional neural networks via attention transfer. arXiv.
53. Tian, Y., Krishnan, D., and Isola, P. (2019). Contrastive representation distillation. arXiv.
54. Shu, C., Liu, Y., Gao, J., Yan, Z., and Shen, C. (2021, January 10–17). Channel-wise knowledge distillation for dense prediction. Proceedings of the IEEE/CVF International Conference on Computer Vision, Montreal, QC, Canada.10.1109/ICCV48922.2021.00526 ↗
55. Heo, B., Kim, J., Yun, S., Park, H., Kwak, N., and Choi, J.Y. (November, January 27). A comprehensive overhaul of feature distillation. Proceedings of the IEEE/CVF International Conference on Computer Vision, Seoul, Republic of Korea.
56. Bai, J., Lu, F., and Zhang, K. (2019). Onnx: Open Neural Network Exchange v1.17.0, Microsoft and Facebook.
57. Jajal, P., Jiang, W., Tewari, A., Kocinare, E., Woo, J., Sarraf, A., Lu, Y.-H., Thiruvathukal, G.K., and Davis, J.C. (2024, January 16–20). Interoperability in deep learning: A user survey and failure analysis of ONNX model converters. Proceedings of the 33rd ACM SIGSOFT International Symposium on Software Testing and Analysis, New York, NY, USA.10.1145/3650212.3680374 ↗
58. Davoodi, P., Gwon, C., Lai, G., and Morris, T. (2019, January 17–21). Tensorrt inference with tensorflow. Proceedings of the GPU Technology Conference, San Jose, CA, USA.
59. Tensorrt-based framework and optimization methodology for deep learning inference on jetson boards10.1145/3508391 ↗