Mesafe Çıkarımlı Görüntü Analizi ile Akıllı Orman Yangını Tespiti

Abstract

Orman yangınları ekosistemler ve insan yerleşimleri için giderek artan bir tehdit oluşturmakta ve hızlı ve doğru tespit sistemleri gerektirmektedir. Bu çalışma, YOLOv5 nesne algılama algoritmasını gelişmiş uzamsal farkındalık için bir mesafe tahmin modülü ile birleştiren yeni bir yangın algılama çerçevesi sunmaktadır. Sistem, gerçek zamanlı görüntü yakalamak için drone'a monte edilmiş kameralardan yararlanarak çeşitli çevresel koşullarda erken yangın tespitine olanak tanır. Modeli eğitmek için 3.058 açıklamalı yangın görüntüsünden oluşan özel bir veri kümesi kullanılmış ve %98'lik bir algılama doğruluğu elde edilmiştir. Sistem, mesafe tahminini entegre ederek hassas yangın lokalizasyonu sağlar ve acil müdahale ekiplerinin müdahaleye öncelik vermesine olanak tanır. LUFFD-YOLO, SWVR-Net ve Faster-RCNN dahil olmak üzere son teknoloji ürünü on yöntemle yapılan karşılaştırmalı deneyler, önerilen yaklaşımın hassasiyet, geri çağırma ve F1-skoru açısından üstünlüğünü göstermiştir. Performans ölçümleri ve eğitim davranışı doğruluk/kayıp eğrileri ve kutu grafikleri aracılığıyla görselleştirilmiştir. Sonuçlar, önerilen sistemin hem sağlamlık hem de güvenilirlik açısından geleneksel yöntemlerden daha iyi performans gösterdiğini ve gerçek zamanlı orman yangını tespiti ve müdahale planlaması için son derece uygun olduğunu doğrulamaktadır. Bu çözüm, mevcut yangın yönetimi altyapılarına potansiyel entegrasyon ile geniş alan dağıtımı için ölçeklenebilir, uygun maliyetli bir alternatif sunmaktadır.

Keywords

• Görüntü Algılama
• Derin Öğrenme
• Yapay Zeka
• Mobil Uygulama

Intelligent Forest Fire Detection Through Image Analysis with Distance Inference

Abstract

Wildfires pose a growing threat to ecosystems and human settlements, requiring rapid and accurate detection systems. This study presents a novel fire detection framework combining the YOLOv5 object detection algorithm with a distance estimation module for enhanced spatial awareness. The system leverages drone-mounted cameras to capture real-time imagery, enabling early fire detection in diverse environmental conditions. A custom dataset of 3,058 annotated fire images was used to train the model, which achieved a detection accuracy of 98%. By integrating distance estimation, the system provides precise fire localization, allowing emergency responders to prioritize intervention. Comparative experiments with ten state-of-the-art methods—including LUFFD-YOLO, SWVR-Net, and Faster-RCNN—demonstrated the superiority of the proposed approach in terms of precision, recall, and F1-score. Performance metrics and training behavior were visualized through accuracy/loss curves and box plots. The results confirm that the proposed system outperforms conventional methods in both robustness and reliability, making it highly suitable for real-time forest fire detection and response planning. This solution offers a scalable, cost-effective alternative for wide-area deployment, with potential integration into existing fire management infrastructures.

Keywords

• Image Detection
• Deep Learning
• Artificial Intelligence
• Mobile Application

Ethical Statement

This article does not require ethics committee approval. This article has no conflicts of interest with any individual or institution.

References

1. CNN Türk, "Orman yangınlarını önleyecek erken tespit sistemi KOZALAK hayata geçirildi," cnnturk.com, Aug. 23, 2022. [Online]. Available: https://www.cnnturk.com/turkiye/orman-yanginlarini-onleyecek-erken-tespit-sistemi-kozalak-hayata-gecirildi. [Accessed: Apr. 21, 2025].
2. Bimetri, "Orman Yangını Erken Uyarı Sistemi," bimetri.com, para. 1, n.d. [Online]. Available: https://www.bimetri.com/cozumler/orman-yangini-erken-uyari-sistemi. [Accessed: Dec. 7, 2022].
3. Habertürk, "Öğrencilerden yangını erken tespit eden drone projesi," haberturk.com, Aug. 4, 2021. [Online]. Available: https://www.haberturk.com/izmir-haberleri/89530192-ogrencilerden-yangini-erken-tespit-eden-drone-projesi. [Accessed: Apr. 21, 2025].
4. Y. Han, B. Duan, R. Guan, G. Yang, and Z. Zhen, "LUFFD-YOLO: A lightweight model for UAV remote sensing forest fire detection based on attention mechanism and multi-level feature fusion," Remote Sensing, vol. 16, no. 12, p. 2177, 2024. doi: 10.3390/rs16122177
5. M. F. S. Titu, M. A. Pavel, G. K. O. Michael, H. Babar, U. Aman, and R. Khan, "Real-Time Fire Detection: Integrating Lightweight Deep Learning Models on Drones with Edge Computing," Drones, vol. 8, no. 9, p. 483, 2024. doi: 10.3390/drones8090483
6. L. Jin, Y. Yu, J. Zhou, D. Bai, H. Lin, and H. Zhou, "SWVR: A lightweight deep learning algorithm for forest fire detection and recognition," Forests, vol. 15, no. 1, p. 204, 2024. doi: 10.3390/f15010204
7. D. Mamadaliev, P. L. M. Touko, J. H. Kim, and S. C. Kim, "ESFD-YOLOv8n: Early smoke and fire detection method based on an improved YOLOv8n model," Fire, vol. 7, no. 9, p. 303, 2024. doi: 10.3390/fire7090303
8. Z. Zhang, L. Tan, and R. L. K. Tiong, "Ship-Fire Net: An improved YOLOv8 algorithm for ship fire detection," Sensors, vol. 24, no. 3, p. 727, 2024. doi: 10.3390/s24030727

9. A. S. Buriboev, K. Rakhmanov, T. Soqiyev, and A. J. Choi, "Improving Fire Detection Accuracy through Enhanced Convolutional Neural Networks and Contour Techniques," Sensors, vol. 24, no. 16, p. 5184, 2024. doi: 10.3390/s24165184
10. P. Vorwerk, J. Kelleter, S. Müller, and U. Krause, "Classification in Early Fire Detection Using Multi-Sensor Nodes—A Transfer Learning Approach," Sensors, vol. 24, no. 5, p. 1428, 2024. doi: 10.3390/s24051428
11. B. Yun, X. Xu, J. Zeng, Z. Lin, J. He, and Q. Dai, "An Improved Unmanned Aerial Vehicle Forest Fire Detection Model Based on YOLOv8," Fire, vol. 8, no. 4, p. 138, 2025. doi: 10.3390/fire8040138
12. S. Muksimova, S. Umirzakova, D. A. Babaraximova, and Y. I. Cho, "Lightweight Fire Detection in Tunnel Environments," Fire, vol. 8, no. 4, p. 134, 2025. doi: 10.3390/fire8040134
13. L. Bu, W. Li, H. Zhang, H. Wang, Q. Tian, and Y. Zhou, "FIRE-YOLOv8s: A Lightweight and Efficient Algorithm for Tunnel Fire Detection," Fire, vol. 8, no. 4, p. 125, 2025. doi: 10.3390/fire8040125
14. L. Deng, S. Wu, J. Zhou, S. Zou, and Q. Liu, "LSKA-YOLOv8n-WIoU: An Enhanced YOLOv8n Method for Early Fire Detection in Airplane Hangars," Fire, vol. 8, no. 2, p. 67, 2025. doi: 10.3390/fire8020067
15. A. W. Ali and S. Kurnaz, "Optimizing Deep Learning Models for Fire Detection, Classification, and Segmentation Using Satellite Images," Fire, vol. 8, no. 2, p. 36, 2025. doi: 10.3390/fire8020036
16. Z. Shi, F. Wu, C. Han, and D. Song, "Research on Fire Detection of Cotton Picker Based on Improved Algorithm," Sensors, vol. 25, no. 2, p. 564, 2025. doi: 10.3390/s25020564
17. U. Ejaz, M. A. Hamza, and H. C. Kim, "Channel Attention for Fire and Smoke Detection: Impact of Augmentation, Color Spaces, and Adversarial Attacks," Sensors, vol. 25, no. 4, p. 1140, 2025. doi: 10.3390/s25041140
18. G. M. I. Alam, N. Tasnia, T. Biswas, M. J. Hossen, S. A. Tanim, and M. S. U. Miah, "Real-Time Detection of Forest Fires Using FireNet-CNN and Explainable AI Techniques," IEEE Access, 2025. doi: 10.1109/ACCESS.2025.3365500
19. T. Maity, A. N. Bhawani, J. Samanta, P. Saha, S. Majumdar, and G. Srivastava, "MLSFDD: Machine Learning-Based Smart Fire Detection Device for Precision Agriculture," IEEE Sensors Journal, vol. 25, no. 3, pp. 2215–2223, 2025. doi: 10.1109/JSEN.2025.3367895
20. A. K. Vishwakarma and M. Deshmukh, "CNNM-FDI: Novel Convolutional Neural Network Model for Fire Detection in Images," IETE Journal of Research, pp. 1–14, 2025. doi: 10.1080/03772063.2025.2324857
21. X. Geng, X. Han, X. Cao, Y. Su, and D. Shu, "YOLOV9-CBM: An improved fire detection algorithm based on YOLOV9," IEEE Access, 2025. doi: 10.1109/ACCESS.2025.3367812
22. J. Liang and J. Cheng, "Mirror Target YOLO: An Improved YOLOv8 Method with Indirect Vision for Heritage Buildings Fire Detection," IEEE Access, 2025. doi: 10.1109/ACCESS.2025.3368213
23. GitHub, "Fire Detection System in Python using HSV Color," github.com, para. 1, May 14, 2022. [Online]. Available: https://github.com/GunarakulanGunaretnam/fire-detection-system-in-python-opencv. [Accessed: Apr. 21, 2025].
24. Y. Chunyu, Z. Yongming, F. Jun, and W. Jinjun, "Texture analysis of smoke for real-time fire detection," in 2009 Second International Workshop on Computer Science and Engineering (WCSE), Oct. 2009, vol. 2, pp. 511-515, Qingdao, China [Online]. Available: https://doi.org/10.1109/WCSE.2009.864. [Accessed: 21 Apr. 2025].
25. P. Barmpoutis, K. Dimitropoulos, K. Kaza, and N. Grammalidis, "Fire detection from images using faster R-CNN and multidimensional texture analysis," in ICASSP 2019 - 2019 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), May 2019, pp. 8301–8305, Brighton, UK [Online]. Available: https://doi.org/10.1109/ICASSP.2019.8682647. [Accessed: 21 Apr. 2025].
26. Z. Xue, H. Lin, and F. Wang, "A small target forest fire detection model based on YOLOv5 improvement," Forests, vol. 13, no. 8, p. 1332, Aug. 2022. doi: 10.3390/f13081332
27. P. Li and W. Zhao, "Image fire detection algorithms based on convolutional neural networks," Case Studies in Thermal Engineering, vol. 19, p. 100625, Jun. 2020. doi: 10.1016/j.csite.2020.100625
28. S. Wu and L. Zhang, "Using popular object detection methods for real time forest fire detection," in 2018 11th International Symposium on Computational Intelligence and Design (ISCID), Dec. 2018, vol. 1, pp. 280–284, Hangzhou, China [Online]. Available: https://doi.org/10.1109/ISCID.2018.00070. [Accessed: 21 Apr. 2025].
29. W. Xiong, "Research on fire detection and image information processing system based on image processing," in 2020 International Conference on Advance in Ambient Computing and Intelligence (ICAACI), Sept. 2020, pp. 106–109, Fuzhou, China [Online]. Available: https://doi.org/10.1109/ICAACI50733.2020.00027. [Accessed: 21 Apr. 2025].
30. N. Li, J. Xue, and H. Li, "An Adaptive Detection Method for Early Smoke of Coal Mine Fire Based on Local Features," in 2022 International Conference on Image Processing and Media Computing (ICIPMC), May 2022, pp. 7–11, Xiamen, China [Online]. Available: https://doi.org/10.1109/ICIPMC55686.2022.00010. [Accessed: 21 Apr. 2025].
31. Kaggle, “Forest Fire Images”, kaggle.com, Jan. 1, 2022. [Online]. Available: https://www.kaggle.com/datasets/mohnishsaiprasad/forest-fire-images. [Accessed: Jan 1, 2024].