A COMPARATIVE STUDY OF DOUBLE-STEP DEEP LEARNING FRAMEWORK FOR BURNED AREA IDENTIFICATION AND SEVERITY ASSESSMENT IN WILDFIRES

Murat Mert Yurdakul; Bülent Bayram; Tolga Bakırman; Hamza Osman İlhan

doi:10.17780/ksujes.1601614

TR EN

YANGINLARDA YANMIŞ ALANLARIN BELİRLENMESİ VE ŞİDDET DEĞERLENDİRMESİ İÇİN ÇİFT ADIMLI DERİN ÖĞRENME ÇERÇEVESİNİN KARŞILAŞTIRMALI BİR ÇALIŞMASI

Öz

Orman yangınlarının daha sık ve yoğun hale gelmesiyle birlikte, doğru tespit ve hasar değerlendirmesi için gelişmiş tekniklerin geliştirilmesi büyük önem taşımaktadır. Bu araştırma, yanmış alanları belirlemek ve yangın şiddetini tahmin etmek için MultiResUNet dahil olmak üzere çeşitli U-Net modellerini kullanan Çift Aşamalı Derin Öğrenme Çerçevesi'ni incelemektedir. Uydu görüntülerinden elde edilen maske çıktıları üzerinde, özellikle 4 ve 5 şiddet seviyelerine odaklanılarak, farklı şiddet seviyelerinin etkileri detaylı bir şekilde incelenmiştir. Ayrıca, Mask R-CNN modeli, önceden eğitilmiş ağırlıklar ve sınırlı spektral girdiler nedeniyle görüntü segmentasyonunda yaşanan zorlukları göstermek için bağımsız olarak incelenmiştir. Yapılan analizler, şiddet aralıklarının granülerliğindeki değişikliklerin model performansını nasıl etkilediğini göstererek, yangın değerlendirmesi için daha ayrıntılı şiddet segmentasyonunun faydalarına dair önemli bilgiler sağlamaktadır. Bu yaklaşım, hasar değerlendirmelerinin doğruluğunu artırma ve yangın yönetimi ile müdahalesinde daha bilinçli kararlar alınmasını destekleme potansiyeline sahiptir.

Anahtar Kelimeler

A COMPARATIVE STUDY OF DOUBLE-STEP DEEP LEARNING FRAMEWORK FOR BURNED AREA IDENTIFICATION AND SEVERITY ASSESSMENT IN WILDFIRES

Abstract

As wildfires become more frequent and intense, it is essential to develop sophisticated techniques for precise detection and damage evaluation. This research examines a Double-Step Deep Learning Framework using several U-Net models, including MultiResUNet, to identify burned areas and estimate severity. Using satellite images, the study explores the effect of different severity levels within mask output, focusing on both 4 and 5 level severity classifications. Additionally, the Mask R-CNN model was evaluated independently for image segmentation, revealing challenges due to its reliance on pretrained weights and limited spectral input. The comparative analysis illustrates how changes in the granularity of severity intervals influence model performance, providing insights into the benefits of more nuanced severity segmentation for wildfire assessment. This approach has the potential to improve the precision of damage assessments and support more informed decision-making in the management and response of wildfires

Keywords

Supporting Institution

TUBITAK

Project Number

122N254

References

Colomba, L., Farasin, A., Monaco, S., Greco, S., Garza, P., Apiletti, D., Baralis, E., & Cerquitelli, T. (2022). A dataset for burned area delineation and severity estimation from satellite imagery. In Proceedings of the 31st ACM International Conference on Information & Knowledge Management (CIKM ’22) (pp. 3893–3897). Association for Computing Machinery. https://doi.org/10.1145/3511808.3557528
Farasin, A., Colomba, L., & Garza, P. (2020). Double-Step U-Net: A deep learning-based approach for the estimation of wildfire damage severity through Sentinel-2 satellite data. Applied Sciences, 10(12), 4332. https://doi.org/10.3390/app10124332
Finney, M. A. (1998). FARSITE: Fire Area Simulator—Model Development and Evaluation (Research Paper RMRS-RP-4, Revised 2004). U.S. Department of Agriculture, Forest Service, Rocky Mountain Research Station.
Han, Y., Zheng, C., Liu, X., Tian, Y., & Dong, Z. (2024). Burned area and burn severity mapping with a transformer-based change detection model. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 17, 13866–13880. https://doi.org/10.1109/JSTARS.2024.3435857
He, K., Gkioxari, G., Dollár, P., & Girshick, R. (2017). Mask R-CNN. Proceedings of the IEEE International Conference on Computer Vision (ICCV), 2961–2969. https://doi.org/10.1109/ICCV.2017.322
He, K., Zhang, X., Ren, S., & Sun, J. (2016). Deep residual learning for image recognition. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 770–778. https://doi.org/10.1109/CVPR.2016.90
Ibtehaz, N., & Rahman, M. S. (2020). Multiresunet: Rethinking the U-Net architecture for multimodal biomedical image segmentation. Neural Networks, 121, 74–87. https://doi.org/10.1016/j.neunet.2019.08.025
Kamal, U., Tonmoy, T. I., Das, S., & Hasan, M. K. (2020). Automatic traffic sign detection and recognition using SegU-Net and a modified Tversky loss function with L1-constraint. IEEE Transactions on Intelligent Transportation Systems, 21(4), 1467–1479. https://doi.org/10.1109/TITS.2019.2911727

Khennou, F., & Akhloufi, M. A. (2023). Improving wildland fire spread prediction using deep U-Nets. Science of Remote Sensing, 8, 100101. https://doi.org/10.1016/j.srs.2023.100101
Li, M., Zhang, Y., Xin, J., Mu, L., Yu, Z., Liu, H., Xie, G., Jiao, S., & Yi, Y. (2021). Early forest fire segmentation based on deep learning. In 2021 CAA Symposium on Fault Detection, Supervision, and Safety for Technical Processes (SAFEPROCESS) (pp. 1–5). IEEE. https://doi.org/10.1109/SAFEPROCESS52771.2021.9693660
Monaco, S., Greco, S., Farasin, A., Colomba, L., Apiletti, D., Garza, P., Cerquitelli, T., & Baralis, E. (2021). Attention to fires: Multi-channel deep learning models for wildfire severity prediction. Applied Sciences, 11(22), 11060. https://doi.org/10.3390/app112211060
Navarro, G., Caballero, I., Silva, G., Parra, P.-C., Vázquez, Á., & Caldeira, R. (2017). Evaluation of forest fire on Madeira Island using Sentinel-2A MSI imagery. International Journal of Applied Earth Observation and Geoinformation, 58, 97–106. https://doi.org/10.1016/j.jag.2017.02.003
Oktay, O., Schlemper, J., Le Folgoc, L., Lee, M., Heinrich, M., Misawa, K., Mori, K., McDonagh, S., Hammerla, N. Y., Kainz, B., Glocker, B., & Rueckert, D. (2018). Attention U-Net: Learning where to look for the pancreas. arXiv preprint, arXiv:1804.03999
Pinto, M. M., Libonati, R., Trigo, R. M., Trigo, I. F., & DaCamara, C. C. (2020). A deep learning approach for mapping and dating burned areas using temporal sequences of satellite images. ISPRS Journal of Photogrammetry and Remote Sensing, 160, 260-274. https://doi.org/10.1016/j.isprsjprs.2019.12.014
Ren, S., He, K., Girshick, R., & Sun, J. (2015). Faster R-CNN: Towards real-time object detection with region proposal networks. In Advances in Neural Information Processing Systems (Vol. 28).
Ronneberger, O., Fischer, P., & Brox, T. (2015). U-Net: Convolutional networks for biomedical image segmentation. In N. Navab, J. Hornegger, W. M. Wells, & A. F. Frangi (Eds.), Medical Image Computing and Computer-Assisted Intervention – MICCAI 2015 (pp. 234–241). Springer International Publishing. https://doi.org/10.1007/978-3-319-24574-4_28
Shamsoshoara, A., Afghah, F., Razi, A., Zheng, L., Fulé, P. Z., & Blasch, E. (2020). Aerial imagery pile burn detection using deep learning: The FLAME dataset. arXiv. https://arxiv.org/abs/2012.14036
Wang, Z., Yang, P., Liang, H., Zheng, C., Yin, J., Tian, Y., & Cui, W. (2022). Semantic Segmentation and Analysis on Sensitive Parameters of Forest Fire Smoke Using Smoke-Unet and Landsat-8 Imagery. Remote Sensing, 14(1), 45. https://doi.org/10.3390/rs14010045
Zou, Y., Sadeghi, M., Liu, Y., Puchko, A., Le, S., Chen, Y., Andela, N., & Gentine, P. (2023). Attention-Based Wildland Fire Spread Modeling Using Fire-Tracking Satellite Observations. Fire, 6(8), 289. https://doi.org/10.3390/fire6080289
Zhou, Z., Rahman Siddiquee, M. M., Tajbakhsh, N., & Liang, J. (2018). Unet++: A nested u-net architecture for medical image segmentation. In Deep Learning in Medical Image Analysis and Multimodal Learning for Clinical Decision Support: 4th International Workshop, DLMIA 2018, and 8th International Workshop, ML-CDS 2018, Held in Conjunction with MICCAI 2018, Granada, Spain, September 20, 2018, Proceedings 4 (pp. 3-11). Springer International Publishing. https://doi.org/10.1007/978-3-030-00889-5_1

Details

Primary Language

English

Subjects

Deep Learning

Journal Section

Research Article

Authors

Murat Mert Yurdakul ^*
0009-0000-7285-7802
Türkiye

Bülent Bayram
0000-0002-4248-116X
Türkiye

Tolga Bakırman
0000-0001-7828-9666
Türkiye

Hamza Osman İlhan
0000-0002-1753-2703
Türkiye

Publication Date

March 3, 2025

Submission Date

December 14, 2024

Acceptance Date

January 26, 2025

Published in Issue

Year 2025 Volume: 28 Number: 1

DOI

https://doi.org/10.17780/ksujes.1601614

IZ

https://izlik.org/JA44NF52TX

Cite

RIS / Bibtex

APA

Yurdakul, M. M., Bayram, B., Bakırman, T., & İlhan, H. O. (2025). A COMPARATIVE STUDY OF DOUBLE-STEP DEEP LEARNING FRAMEWORK FOR BURNED AREA IDENTIFICATION AND SEVERITY ASSESSMENT IN WILDFIRES. Kahramanmaraş Sütçü İmam Üniversitesi Mühendislik Bilimleri Dergisi, 28(1), 513-523. https://doi.org/10.17780/ksujes.1601614