Aktarımlı Öğrenme ile SENTINEL-2 Görüntülerinden Kıyı Çizgisi Bölütlemesi

Yıl 2021, Cilt: 3 Sayı: 1, 1 - 7, 15.06.2021

Selennur Karagöl , Bülent Bayram , Fırat Erdem , Tolga Bakirman

https://doi.org/10.51489/tuzal.830052

Öz

Kıyı çizgileri küresel ısınma, nüfus artışı, çevre kirliliği, kentleşme etkileriyle sürekli değişir. Doğal ve antropojenik etkilerle meydana gelen değişikleri tespit etmek için kıyı alanlarının izlenmesi gerekmektedir. Kıyı alanlarındaki değişimlerin sürdürülebilir bir şekilde izlenmesi, kıyı kaynak yönetimi, çevresel koruma ve planlama açısından oldukça önemli rol oynamaktadır. Uydu görüntüleri bu amaç için doğru, güvenilir, zamansal ve güncel bilgiler sağlamaktadır. Derin öğrenme(DL) ve aktarımlı öğrenme(TL) yaklaşımları kıyı çizgisi çıkartılmasında yeni olanaklar sağlamaktadır. Sunulan çalışmada, SENTINEL-2 görüntülerinden aktarımlı öğrenmeye dayalı, U-NET mimarisi kullanılarak, bir kara ve su bölütlemesi yaklaşımı önerilmiştir. Önceden eğitilmiş modele ait özellikler ve ağırlıklar için, LANDSAT-8 görüntüleri ile gerçekleştirilen derin öğrenme çalışmasından yararlanılmıştır. U-Net mimarisi kullanılan ağda, mavi, kırmızı ve yakın kızıl ötesi bantlarından oluşan tam çerçeve SENTINEL-2 görüntülerinden 8’i eğitim, 7’si test aşamasında kullanılmıştır. Tam çerçeve görüntüler 512x512 boyutlarında kırpılarak eğitim ve test için sırasıyla 115 ve 235 görüntü parçası oluşturulmuştur. Ortalama doğruluk, duyarlılık, hassasiyet, özgünlük ve F-skor değerleri sırasıyla 0.9917, 0.9927, 0.9908, 0.9907 ve 0.9917 olarak hesaplanmıştır. Çalışmanın sonuçlarına göre, aktarımlı öğrenme kullanılarak az miktarda görüntü ile yüksek doğruluklu kıyı çizgisi elde etmek mümkündür.

Anahtar Kelimeler

Derin öğrenme, Aktarımlı Öğrenme, Kıyı Çizgisi Çıkartma, SENTINEL-2, LANDSAT-8

Shoreline Segmentation from SENTINEL-2 Imagery by Transfer Learning

Öz

Global warming, increasing population, environmental pollution and urbanization can constantly affect coastal areas. Therefore, sustainable monitoring of coastal zones is vital to detect changes which can occur due to natural and anthropogenic effects. Thus, sustainable shoreline monitoring is essential for coastal resource management, environmental protection and planning. Satellite images provide accurate, reliable, temporal and up-to-date information for this purpose. State-of-the-art deep learning (DL) and transfer learning approaches brought new opportunities for shoreline extraction. In this study, a transfer learning based water-body segmentation framework with U-Net architecture from SENTINEL-2 imagery has been proposed. The pre-trained weights have been obtained from another study which is a network trained with LANDSAT-8 imageries. The training of used U-Net architecture was carried out using SENTINEL-2 imagery which consists of blue, red and NIR bands with 8 and 7 full frames for training and testing, respectively. Images have been cropped as 512x512 pixels and 115 and 235 patches have been created for the training and testing dataset, respectively. Average accuracy, recall, precision, specivity and F-score of the model values has been calculated as 0.9917, 0.9927, 0.9908, 0.9907 and 0.9917, respectively. The results show that it is possible to obtain shoreline with high accuracy with limited data using transfer learning.

Anahtar Kelimeler

Deep Learning, Transfer Learning, Shoreline Extraction, SENTINEL-2, LANDSAT-8

Kaynakça

• Alesheikh, A. A., Ghorbanalı, A. & Nouri, N. (2007). Coastline change detection using remote sensing. International Journal of Environmental Science & Technology, 4(1), 61-66.
• Bayram, B., Erdem, F., Akpınar, B., Ince, A. K., Bozkurt, S., Reis, H. C. & Seker, D. Z. (2017). The Efficiency of Random Forest Method for Shoreline Extraction from LANDSAT-8 and GOKTURK-2 Imageries. ISPRS Annals of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Volume IV-4/W4, 141-145.
• Choung, , Y. J., & Jo, M. H. (2017). Comparison between a Machine-Learning-Based Method and a Water-Index-Based Method for Shoreline Mapping Using a High-Resolution Satellite Image Acquired in Hwado Island, South Korea. Journal of Sensors, vol. 2017, Article ID 8245204.
• Dixon, B. & Candade, N. (2008). Multispectral landuse classification using neural networks and support vector machines: one or the other, or both?. International Journal of Remote Sensing, 29(4), 1185-1206. Erdem, F., Bayram, B., Bakirman, T., Bayrak, O.C. & Akpinar, B. (2020). An Ensemble Deep Learning Based Shoreline Segmentation Approach (WaterNet) from Landsat 8 OLI images, Advances in Space Research, doi: https://doi.org/10.1016/j.asr.2020.10.043.
• Gens, R. (2010). Remote sensing of coastlines: detection, extraction and monitoring. International Journal of Remote Sensing, 31(7), 1819–1836.
• Goodfellow, I., Bengio, Y., Courville, A. & Bengio, Y. (2016). Deep Learning, MIT Press, Cambridge.
• Guariglia, A., Buonamassa, A., Losurdo, A., Saladino, R., Trivigno, M. L., Zaccagnino, A., & Colangelo, A. (2006). A multisource approach for coastline mapping & identification of the shoreline changes. Annals of Geophysics, 49(1), 295–304.
• Işıkdoğan, F., Bovik, A. C. & Passalacqua, P. (2017). Surface Water Mapping by Deep Learning. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 10(11), 4909-4918.
• İncekara, A. H., Seker, D. Z. & Bayram, B. (2018). Qualifying the LIDAR-Derived Intensity Image as an Infrared Band in NDWI-Based Shoreline Extraction. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 11(12), 5053-5062.
• Kalkan, K., Bayram, B., Maktav, D. & Sunar, F. (2013). Comparison of support vector machine and object based Classification methods for coastline detection. International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Volume XL-7/W2, 125-127.
• Kaur, T. & Gandhi, T. K. (2019). Deep convolutional neural networks with transfer learning for automated brain image classification. Machine Vision and Applications 31,1-16.
• Lary, D. J., Alavi, A. H., Gandomi, A. H. & Walker, A. L. (2016). Machine learning in geosciences and remote sensing. Geoscience Frontiers, 7(1), 3-10.
• Li, R., Liu, W., Yang, L., Sun, S., Hu, W., Zhang, F. & Li, W. (2018). DeepUNet: a deep fully convolutional network for pixel-level sea-land segmentation. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 11(11), 3954-3962.
• Moore, L. (2000). Shoreline Mapping Techniques. Journal of Coastal Research, 16(1), 111-124. Retrieved November 18, 2020, from http://www.jstor.org/stable/4300016.
• Nazerdeylami, A., Majidi, A., & Movaghar, A. (2019). Smart Coastline Environment Management Using Deep Detection of Manmade Pollution and Hazards. 2019 5th Conference on Knowledge Based Engineering and Innovation (KBEI), Tehran, Iran, 2019, pp. 332-337.
• Pardo-Pascual, J. E., Almonacid-Caballer, J., Ruiz, L. A. & Palomar-Vazquez, J. (2012). Automatic extraction of shorelines from LANDSAT TM and ETM+ multi-temporal images with subpixel precision. Remote Sensing of Environment, 123, 1-11.
• Pardo-Pascual, J., Sánchez-García, E., Almonacid-Caballer, J., Palomar, J., Priego, J. Fernández-Sarría, A & Balaguer-Beser, A.. (2018). Assessing the Accuracy of Automatically Extracted Shorelines on Microtidal Beaches from LANDSAT 7, LANDSAT 8 and SENTINEL-2 Imagery. Remote Sensing. 10. 326. 10.3390/rs10020326.
• Patterson, J., & Gibson, A. (2017). Deep Learning: A Practitioner's Approach, First Edition, O'Reilly Media, California.
• Ronneberger, O., Fischer, P., & Brox, T. (2015). U-net: Convolutional networks for biomedical image segmentation. In International Conference on Medical image computing and computerassisted intervention (pp. 234-241). Springer, Cham.
• Song, S., Liu, J., Liu, Y., Feng, G., Han, H., Yao, Y., & Du, M. (2020). Intelligent Object Recognition of Urban Water Bodies Based on Deep Learning for Multi-Source and Multi-Temporal High Spatial Resolution Remote Sensing Imagery. Sensors, 20(2), 397.
• Sreekesh, S., Kaur, N. & Sreerama Naik, S.R. (2020). An OBIA and Rule Algorithm for Coastline Extraction from High- and Medium-Resolution Multispectral Remote Sensing Images. Remote Sens Earth Syst Sci 3, 24–34.
• Syrris, V., Hasenohr, P., Delipetrev, B., Kotsev, A., Kempeneers, P. & Soille, P. (2019). Evaluation of the Potential of Convolutional Neural Networks and Random Forests for Multi-Class Segmentation of SENTINEL-2 Imagery. Remote Sensing. 11. 907. 10.3390/rs11080907.
• Torrey, L., & Shavlik, J. (2009). Chapter 11 Transfer Learning.
• URL-1: https://SENTINEL.esa.int/web/SENTINEL/user-guides/SENTINEL-2-msi/resolutions/spatial[Erişim Tarihi: 21.11.2020]
• URL-2: https://osmdata.openstreetmap.de/data/water-polygons.html[Erişim Tarihi: 21.11.2020]
• URL-3: https://github.com/tensorflow/tensorflow[Erişim Tarihi: 21.11.2020]
• URL-4: https://github.com/keras-team/keras[Erişim Tarihi: 21.11.2020]
• Wieland, M., & Martinis, S. & Li, Y. (2019). Semantic segmentation of water bodies in multi-spectral satellite images for situational awareness in emergency response. ISPRS - International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences. XLII-2/W16. 273-277. 10.5194/isprs-archives-XLII-2-W16-273-2019.
• Xu, Y., Wu, L., Xie, Z. & Chen, Z. (2018). Building Extraction in Very High Resolution Remote Sensing Imagery Using Deep Learning and Guided Filters. Remote Sensing, 10(1), 144-161.
• Yang, L., Tian, S., Yu, L., Ye, F., Qian, J. & Qian, Y. (2015). Deep learning for extracting water body from LANDSAT imagery. International Journal of Innovative Computing, Information and Control, 11(6), 1913-1929.
• Yousef A., & Iftekharuddin K. (2014). Shoreline extraction from the fusion of LiDAR DEM data and aerial images using mutual information and genetic algorithms. International Joint Conference on Neural Networks (IJCNN) (pp. 1007–1014). Beijing, China.
• Yu S., Mou Y., Xu d., You X., Zhou L., & Zeng, W. (2013). A New Algorithm for Shoreline Extraction from Satellite Imagery with Non-Separable Wavelet and Level Set Method. International Journal of Machine Learning and Computing, 3(1), 158-163.
• Yu, L., Wang, Z., Tian, S., Ye, F., Dıng, J. & Kong, J. (2017). Convolutional Neural Networks for WaterBody Extraction from Landsat Imagery. International Journal of Computational Intelligence and Applications, 16(1), 1750001.
• Zhang, Y., Li, X., Zhang, J. & Song, D. (2013). A Study on Coastline Extraction and Its Trend Based on Remote Sensing Image Data Mining. Abstract and Applied Analysis, vol. 2013, Article ID 693194, 6 pages, https://doi.org/10.1155/2013/693194.
• Zhang, Y., Li, X., Zhang, J. & Song, D. (2013). A Study on Coastline Extraction and Its Trend Based on Remote Sensing Image Data Mining. Abstract and Applied Analysis, vol. 2013, Article ID 693194, 6 pages, https://doi.org/10.1155/2013/693194.
• Zheng, G., Peng, L., Tao, G. & Wang, C. (2011). Remote sensing analysis of Bohai Bay West Coast shoreline changes. In Proceedings IEEE International Conference on Spatial Data Mining and Geographical Knowledge Services, pp. 549–552.