Real-Time Forest Fire Detection and Spread Prediction Using Transformer-Based Vision Models on Edge Computing Nodes: A Case Study of the Aegean Pine Forest Zone, Greece
DOI:
https://doi.org/10.68099/asnj.2025.86Keywords:
Forest Fire Detection, Spread Prediction, Cellular AutomatonAbstract
Increasing forest fire frequency in Eastern Mediterranean Pinus brutia ecosystems demands rapid automated detection and real-time spread prediction. This paper presents EgeoFireNet, an end-to-end edge-computing system deployed on NVIDIA Jetson Orin NX nodes within a 43-camera dual-channel monitoring network in Chios, Greece. The system integrates a fine-tuned Swin Transformer detection model with a physics-informed cellular automaton spread prediction module. Optimized via TensorRT INT8, the detection model processes 22.4 FPS, achieving a 0.889 [email protected] and a 0.912 F1-score for fire and smoke. Initialized by real-time segmentation masks and sensor data, the prediction module generates 15-minute georeferenced forecasts in 4.2 seconds with a 0.76 Sorensen-Dice coefficient. EgeoFireNet reduces median end-to-end alert latency to 87 seconds a 12-to-24-fold improvement over conventional helicopter patrols while maintaining an operational false positive rate of 0.023 alerts/camera/hour.
References
[1] European Forest Fire Information System (EFFIS), Annual Fire Danger Season Analysis 2023. Ispra: European Commission Joint Research Centre, 2024.
[2] A. Mohapatra and T. Trinh, "Early wildfire detection technologies in practice A review," Sustainability, vol. 14, no. 19, pp. 12270, Sep. 2022. DOI: https://doi.org/10.3390/su141912270
[3] T. Nawaz, O. Maguire, K. G. Clarke, R. Holsgrove, A. Al-Mamoon, and R. Bhakta, "Forest fire detection using distilled vision transformers for deployment on edge devices," preprint, Staffordshire University, 2024.
[4] Z. Liu, Y. Lin, Y. Cao, H. Hu, Y. Wei, Z. Zhang, S. Lin, and B. Guo, "Swin transformer: Hierarchical vision transformer using shifted windows," in Proc. ICCV, pp. 10012-10022, 2021. DOI: https://doi.org/10.1109/ICCV48922.2021.00986
[5] B. H. Hopkins, L. O. O’Neill, F. A. Afghah, A. R. Razi, A. W. Watts, P. F. Fule, and J. C. Coen, "FLAME 2: Fire detection and modeLing: Aerial multi-spectral imagE dataset," IEEE DataPort, 2022.
[6] R. Ghali and M. A. Akhloufi, "Deep learning approaches for wildland fires remote sensing: Classification, detection, and segmentation," Remote Sens., vol. 15, no. 7, pp. 1821, Mar. 2023. DOI: https://doi.org/10.3390/rs15071821
[7] R. C. Rothermel, A Mathematical Model for Predicting Fire Spread in Wildland Fuels. Ogden: USDA Forest Service Research Paper INT-115, 1972. DOI: https://doi.org/10.2737/INT-RP-115
[8] L. Mu, Y. Yang, B. Wang, Y. Zhang, N. Feng, and X. Xie, "Edge computing-based real-time forest fire detection using UAV thermal and color images," IEEE J. Sel. Top. Appl. Earth Obs. Remote Sens., vol. 18, pp. 6760 - 6771, 2025. DOI: https://doi.org/10.1109/JSTARS.2025.3528652
[9] G. Coleman, W. Salter, and M. Walsh, "OpenWeedLocator (OWL): An open-source, low-cost device for fallow weed detection," Sci. Rep., vol. 12, no. 1, pp. 170, Jan. 2022. DOI: https://doi.org/10.1038/s41598-021-03858-9
[10] A. Stergiopoulos, G. Botzoris, and C. Basbas, "Optimal placement of forest fire detection cameras using geographic information systems and analytic hierarchy process," Appl. Sci., vol. 13, no. 6, pp. 3563, Mar. 2023.
[11] J. Li, R. Xu, and Y. Liu, "An improved forest fire and smoke detection model based on YOLOv5," Forests, vol. 14, no. 5, pp. 833, May 2023. DOI: https://doi.org/10.3390/f14040833
[12] K. Muhammad, J. Ahmad, Z. Lv, P. Bellavista, P. Yang, and S. W. Baik, "Efficient deep CNN-based fire detection and localization in video surveillance applications," IEEE Trans. Syst. Man Cybern. Syst., vol. 49, no. 7, pp. 1419-1434, Jul. 2019. DOI: https://doi.org/10.1109/TSMC.2018.2830099
[13] A. Dosovitskiy, L. Beyer, A. Kolesnikov, D. Weissenborn, X. Zhai, T. Unterthiner, M. Dehghani, M. Minderer, G. Heigold, S. Gelly, J. Uszkoreit, and N. Houlsby, "An image is worth 16x16 words: Transformers for image recognition at scale," in Proc. ICLR, 2021.
[14] L. Papa, P. Russo, I. Amerini, and L. Zhou, "A survey on efficient vision transformers: Algorithms, techniques and performance benchmarking," IEEE Trans. Pattern Anal. Mach. Intell., pp. 1-20, 2024.
[15] A. Khan, B. Hassan, S. Khan, R. Ahmed, and A. Abuassba, "DeepFire: A novel dataset and deep transfer learning benchmark for forest fire detection," Mob. Inf. Syst., vol. 2022, pp. 5358359, 2022.
[16] L. Papa, P. Russo, I. Amerini, and L. Zhou, "A survey on efficient vision transformers: Algorithms, techniques, and performance benchmarking," IEEE Trans. Pattern Anal. Mach. Intell., pp. 1-20, 2024. DOI: https://doi.org/10.1109/TPAMI.2024.3392941
[17] A. Khan, B. Hassan, S. Khan, R. Ahmed, and A. Abuassba, "DeepFire: A novel dataset and deep transfer learning benchmark for forest fire detection," Mob. Inf. Syst., vol. 2022, pp. 5358359, 2022. DOI: https://doi.org/10.1155/2022/5358359
Downloads
Published
Issue
Section
License
Copyright (c) 2025 Elif Yildiz (Author)
This work is licensed under a Creative Commons Attribution 4.0 International License.