Agricultural Crop Detection with a Machine Learning Algorithm from Fused Sentinel-1 SAR and Landsat-8 Optical Data

Yıl 2022, , 1 - 19, 14.03.2022

Müslüm Altun , Mustafa Türker

https://doi.org/10.48123/rsgis.999749

Öz

In this study, the effect of fused single-date Sentinel-1 SAR and Landsat-8 data on agricultural crop detection by classification was investigated. A region in Manisa, Turkey was chosen as study area. Ehlers method was used for image fusion and random forest (RF) machine learning algorithm was used for image classification. Classification was carried out using only Sentinel-1 data, only Landsat-8 data, and the fused Sentinel-1 and Landsat-8 datasets. In classification using only Sentinel-1 VV or VH band, the overall accuracies were calculated at the level of 35%. The combined use of Sentinel-1 VV and VH bands contributed about 6% to classification performance. The accuracy (71,18%) calculated from classification performed using Landsat-8 data alone is quite high when compared to classification performed using Sentinel-1 data alone. In classification with the fused Landsat-8 and Sentinel-1 DD dataset and with the fused Landsat-8 and Sentinel-1 DY dataset, the accuracies were calculated as 80.44% and 82.16%, respectively. The highest accuracy (87.72%) was obtained in classification performed using the fused Landsat-8 and Sentinel-1 VV+VH bands dataset. Based on the results, the use of the fused single-date Landsat-8 and Sentinel-1 VV+ VH bands dataset in classification was found to significantly increase the accuracy.

Anahtar Kelimeler

Sentinel-1 SAR, Landsat 8, Classification, Random Forest (RF), Ehlers Image Fusion

Kaynaştırılmış Sentinel-1 SAR ve Landsat-8 Optik Veriden Makine Öğrenme Algoritması ile Tarımsal Ürün Tespiti

Öz

Bu çalışmada, tek tarihe ait kaynaştırılmış Sentinel-1 Yapay Açıklıklı Radar (Synthetic Aperture Radar-SAR) ve Landsat-8 verilerinin sınıflandırma ile tarımsal ürün tespitine olan etkisi araştırılmıştır. Çalışma alanı olarak, Manisa’da bir bölge seçilmiştir. Görüntü kaynaştırma için Ehlers yöntemi, görüntü sınıflandırma için rastgele orman (RO) makine öğrenme algoritması kullanılmıştır. Sınıflandırma orjinal Sentinel-1 verisi ile orjinal Landsat-8 verisi ile ve kaynaştırılmış veri setleri ile gerçekleştirilmiştir. Orjinal Sentinel-1 DD veya DY bandı ile yapılan sınıflandırmanın genel doğruluğu %35 mertebesinde hesaplanmıştır. Sentinel-1 DD ve DY bantların birlikte kullanılmasının sınıflandırma performansına katkısı %6 kadar olmuştur. Orjinal Landsat-8 verisi ile yapılan sınıflandırma sonucu hesaplanan genel doğruluk değeri (%71,18), orjinal Sentinel-1 verisine göre oldukça yüksektir. Landsat-8 ile kaynaştırılmış Sentinel-1 DY ve DD batları veri setleri ile yapılan sınıflandırmanın genel doğruluğu sırasıyla %80,44 ve %82,16 olarak hesaplanmıştır. En yüksek genel doğruluk değeri (%87,72), Landsat-8 ile kaynaştırılmış Sentinel-1 DD+DY bantları veri seti ile yapılan sınıflandırmada elde edilmiştir. Elde edilen bulgulara göre, sınıflandırmada kaynaştırılmış tek tarihli Landsat-8 ve Sentinel-1 DD+DY bantları veri setinin kullanılması doğruluğu önemli oranda artırmaktadır.

Anahtar Kelimeler

Sentinel-1 SAR, Landsat-8, Sınıflandırma, Rastgele Orman (RO), Ehlers Görüntü Kaynaştırma

Kaynakça

• Adrian, J., Vasit Sagan, V., & Maimaitijiang, M. (2021). Sentinel SAR-optical fusion for crop type mapping using deep learning and Google Earth Engine. ISPRS Journal of Photogrammetry and Remote Sensing, 175(2021), 215-235. doi: 10.1016/j.isprsjprs.2021.02.018.
• Akar, Ö. (2013). Rastgele orman sınıflandırıcısına doku özellikleri entegre edilerek benzer spektral özellikteki tarımsal ürünlerin sınıflandırılması (Doktora tezi), Karadeniz Teknik Üniversitesi, Fen Bilimleri Enstitüsü, Türkiye.
• Alparone, L., Alazzi, B., Baronti, S., Garzelli, A., Nencini, F. & Selva, M. (2008). Multispectral and panchromatic data fusion assessment without reference. Photogrammetric Engineeering and Remote Sensing, 74(2), 193-200.
• Alparone, L., Baronti, S., Garzelli, A., & Nencini, F. (2004). A global quality measurement of pan-sharpened multispectral imagery. IEEE Geoscience and Remote Sensing Letters, 1(4), 313-317. doi: 10.1109/LGRS.2004.836784.
• ArcGIS User Guide. (2021, Nisan 22). ArcGIS Desktop User Guide Documentation, Retrieved from https://desktop.arcgis.com/en/documentation/
• Archer, K. J., & Kimes, R. V. (2008). Emprical characterization of random forest variable importance measure. Computational Statistics & Data Analysis, 52(4), 2249-2260. doi: 10.1016/j.csda.2007.08.015.
• Breiman, L. (2001). Random forests, Machine Learning, 45(1), 5-32.
• Breiman, L. (2003). Manual setting up, using, and understanding random forests. RColorBrewer MASS, 4(0), 1-33.
• Breiman, L., Friedman, J. H., Olshen, R. A., & Stone C. J., (1984). Classification and Regression Trees. Boca Raton, Chapman & Hall/CRC Press.
• Brisco, B., & Brown, R. J. (1995). Multidate SAR/TM synergism for crop classification in Western Canada. Photogrammetric Engineering & Remote Sensing, 61(8), 1009-1014.
• Bush, T. F., & Ulaby, F. T. (1978). An evaluation of radar as a crop classifier. Remote Sensing of Environment, 7(1), 15-36.
• Campbell, J.B., Wynne, R.H. (1996). Introduction to Remote Sensing. New York, London, Guilford Press.
• Cao, J., Cai, X., Tan, J., Cui, Y., Xie, H., Liu, F., Yang, L., & Luo, Y. (2020). Mapping paddy rice using Landsat time series data in the Ganfu Plain irrigation system, Southern China, from 1988-2017. International Journal of Remote Sensing, 42(4), 1556-1576. doi: 10.1080/01431161.2020.1841321.
• Chen, S., Useya, J., & Hillary Mugiyo, H. (2020). Decision-level fusion of Sentinel-1 SAR and Landsat 8 OLI texture features for crop discrimination and classification: case of Masvingo, Zimbabwe. Heliyon, 6(11), 1-14.
• Cohen, J. (1960). A coefficient of agreement for nominal scales. Educational and Psychological Measurement, 20(1), 37-46.
• Congalton, R. G. (1991). A review of assessing the accuracy of classifications of remotely sensed data. Remote Sensing of Environment, 37(1), 35-46.
• Congalton, R. G., Green, K. (2008). Assessing the accuracy of remotely sensed data: Principles and practices. Boca Raton, CRC Press.
• Copernicus. (2021, Mart 30). Copernicus Open Access Hub. Retrieved from https://scihub.copernicus.eu/dhus/#/home
• Csillik, O., & Belgiu, M. (2017, May). Cropland mapping from Sentinel-2 time series data using object-based image analysis. In Proceedings of the 20th AGILE International Conference on Geographic Information Science Societal Geo-Innovation Celebrating, Wageningen, The Netherlands (pp. 9-12).
• Dammavalam, S. R., Maddala, S., & Mhm, K. P. (2012). Quality assessment of pixel-level image fusion using fuzzy logic. International Journal on Soft Computing, 3(1), 11-23. doi:10.5121/ijsc.2012.3102.
• Dimov, D., Kuhn, J., & Conrad, C. (2016, July). Assessment of cropping system diversity in the fergana valley through image fusion of Landsat 8 and Sentinel-1. In 23rd ISPRS Congress, 2016. (pp.173-180). XXIII ISPRS.
• EarthExplorer. (2021, Mart 30). USGS Earth Explorer. Retrieved from https://earthexplorer.usgs.gov
• Erdas Imagine User Guide. (2021, Nisan 22). Hexagon Erdas Imagine User Guide, Retrieved from https://www.hexagongeospatial.com/
• ESA Copernicus Open Access Hub. (2021, Mart 30). Overview, Sentinel-1 Data Offer, Retrieved from https://scihub.copernicus.eu/userguide/
• Filipponi, F. (2019). Sentinel-1 GRD preprocessing workflow. Multidisciplinary Digital Publishing Institute Proceedings, 18(1), 11-15. doi:10.3390/ECRS-3-06201.
• Foody, G. M., M. B. Mcculloch, M. B., & W. B. Yates, W.B. (1994). Crop classification from C-band polarimetric radar data. International Journal of Remote Sensing, 15(14), 2871-2885. doi: 10.1080/01431169408954289.
• Forget, Y., Shimoni, M., Gilbert, M., & Linard, C. (2018). Complementarity between Sentinel-1 and Landsat 8 imagery for built-up mapping in Sub-Saharan Africa. Preprints, doi: 10.20944/preprints201810.0695.v1.
• Gumma, M. K., Nelson, A., Thenkabail, P. S., & Singh, A. N. (2011). Mapping rice areas of South Asia using MODIS multitemporal data. Journal of Applied Remote Sensing, 5(1), 1-26. doi: 10.1117/1.3619838.
• Gungor, O. (2008). Multi sensor multi resolution image fusion (Doctoral dissertation), Purdue University, USA.
• Horning, N. (2010, December). Random forests: An algorithm for image classification and generation of continuous fields data sets. In International Conference on Geoinformatics for Spatial Infrastructure Development in Earth and Allied Sciences, 2010. Proceedings. GIS-IDEAS.
• Hütt, C., Koppe, W., Miao, Y., & Bareth, G. (2016). Best accuracy land use/land cover (LULC) classification to derive crop types using multitemporal, multisensor, and multi-polarization SAR satellite images. Remote Sensing, 8(8), 684-698. doi:10.3390/rs8080684.
• Klonus, S., & Ehlers, M. (2009, July). Performance of evaluation methods in image fusion. In 2009 12th International Conference on Information Fusion (pp. 1409-1416). IEEE.
• Kussul, N., Lavreniuk, M., Skakun, S., & Shelestov, A. (2017, May). Deep learning classification of land cover and crop types using remote sensing data. IEEE Geoscience and Remote Sensing Letters, 14(5), 778-782. doi: 10.1109/LGRS.2017.2681128.
• Lee, J.S., Jurkevich, I., Dewaele, P., Wambacq, P., & Oosterlinck, A. (1994). Speckle filtering of synthetic aperture radar images: A review. Remote Sensing Reviews, 8(4), 313-340. doi: 10.1080/02757259409532206.
• Lemoine, G., & Leo, O. (2015, November). Crop mapping applications at scale: Using google earth engine to enable global crop area and status monitoring using free and open data sources. In International Geoscience and Remote Sensing Symposium, (IGARSS) (pp. 1496-1499). IEEE. doi: 10.1109/IGARSS.2015.7326063.
• Liaw, A., & Wiener, M. (2002). Classification and regression by random forest. R News, 2(3), 18-22.
• Liu, M.W., Ozdogan, M., & Zhu, X. (2014). Crop type classification by simultaneous use of satellite images of different resolutions. IEEE Transactions on Geoscience and Remote Sensing, 52(6), 3637-3649. doi: 10.1109/TGRS.2013.2274431.
• Lussem, U., Hütt, C., & Waldhoff, G. (2016, July). Combined analysis of Sentinel-1 and Rapid Eye data for improved crop type classification: An early season approach for rapeseed and cereals. International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, 41(2016), 959-963. doi:10.5194/isprsarchives-XLI-B8-959-2016.
• Mather, P., & Tso, B. (2009). Classification methods for remotely sensed data. BW: CRC Press.
• MathWorks. (2021, Mayıs 8). MathWorks Makers of MATLAB and Simulink. Retrieved from https://www.mathworks.com/
• McNairn H., & Shang J. (2016). A Review of multitemporal synthetic aperture radar (SAR) for crop monitoring. In Ban, Y. (Eds.) Multitemporal remote sensing. Remote sensing and digital image processing (Vol. 20, pp. 217-340). Canada, Ottowa, ON: Springer Press.
• McNairn, H., & Shang, J. (2016). A review of multitemporal synthetic aperture radar (SAR) for crop monitoring. Multitemporal remote sensing, 317-340.
• McNairn, H., & Shang, J. (2016). A review of multitemporal synthetic aperture radar (SAR) for crop monitoring. Multitemporal remote sensing, 20(2016), 317-340.
• Nasirzadehdizaji, R., Sanli, F. B., Cakir, Z., & Sertel, E. (2019, July). Crop mapping improvement by combination of optical and SAR datasets. In 2019 8th International Conference on Agro Geoinformatics, (pp. 1-6). IEEE. doi: 10.1109/Agro-Geoinformatics.2019.8820604.
• Nuthammachot, N., & Stratoulias, D. (2019). Fusion of Sentinel-1A and Landsat-8 images for improving land use/land cover classification in Songkla Province, Thailand. Applied Ecology and Environmental Research, 17(2), 3123-3135. doi: 10.15666/aeer/1702_31233135.
• Otukei, J. R., Blaschke, T., & Collins, M. (2015). Fusion of TerraSAR-x and Landsat ETM+ data for protected area mapping in Uganda. International Journal of Applied Earth Observation and Geoinformation, 38(2015), 99-104. doi: 10.1016/j.jag.2014.12.012.
• Ozdarici A. & Turker, M. (2006, July). Field-based classification of agricultural crops using multi-scale images, Proceedings of the 1st International Conference on Object-based Image Analysis, 2006. (pp. 1-7). OBIA’06.
• Pal, M. (2005a, February). Multiclass approaches for support vector machine based land cover classification, In Proceedings of Map India, 8th Annual International Conference and Exhibition in the Field of GIS, GPS, Aerial Photography and Remote Sensing, 2005.
• Pal, M. (2005). Random forest classifier for remote sensing classification. International Journal of Remote Sensing, 26(1), 217-222. doi: 10.1080/01431160412331269698.
• Pal, M., & Mather, P. M. (2003). An assessment of the effectiveness of decision tree methods for land cover classification. Remote Sensing of Environment, 86(4), 554-565. doi:10.1016/S0034-4257(03)00132-9.
• Rodriguez-Galiano, V. F., Ghimire, B., Rogan, J., Chica-Olmo, M., & Rigol-Sanchez, J. P. (2012). An assessment of the effectiveness of a random forest classifier for land cover classification. ISPRS Journal of Photogrammetry and Remote Sensing, 67(2012), 93-104. doi:10.1016/j.isprsjprs.2011.11.002.
• Schowengerdt, R. A. (2007). Remote sensing: Models and methods for image processing. AP: Elseiver Press.
• Siachalou, S., Mallinis, G., & Tsakiri-Strati, M. (2015). A hidden markov models approach for crop classification: Linking crop phenology to time series of multisensor remote sensing data. Remote Sensing, 7(4), 3633-3650. doi:10.3390/rs70403633.
• Skriver, H. (2012). Crop classification by multitemporal C- and L-band single- and dual polarization and fully polarimetric SAR. IEEE Transactions on Geoscience and Remote Sensing, 50(6), 2138-2149. doi: 10.1109/TGRS.2011.2172994.
• Sonobe, R., Tani, H., Wang, X., Kobayashi, N., & Shimamura, H. (2014). Random forest classification of crop type using multi-temporal TerraSAR-X dual-polarimetric data. Remote Sensing Letters, 5(2), 157-164. doi: 10.1080/2150704X.2014.889863.
• Suwarsono, N., Prasasti, I., Nugroho, J. T., Sitorus, J., & Triyono, D. (2018). Detectıng the lava flow deposıts from 2018 anak krakatau eruptıon usıng data fusıon Landsat-8 optıc and Sentınel-1 SAR. International Journal of Remote Sensing and Earth Sciences, 15(2), 157-166. doi: 10.30536/j.ijreses.2018.v15.a3078.
• T.C. Tarım ve Orman Bakanlığı. (2021, Mart 30). T.C. Tarım ve Orman Bakanlığı Anasayfası, Retrieved from https://www.tarimorman.gov.tr/
• Thenkabail, P. S., Hanjra, M. A., Dheeravath, V., & Gumma, M. (2010). A holistic view of global croplands and their water use for ensuring global food security in the 21st century through advanced remote sensing and non-remote sensing approaches. Remote Sensing, 2(1), 211-261. doi: 10.3390/rs2010211.
• Thenkabail, P. S., Hanjra, M. A., Dheeravath, V., & Gumma, M. (2011). Global croplands and their water use remote sensing and non-remote sensing perspectives. Weng Q (Ed.), Advances in Environmental Remote Sensing: Sensors, Algorithms, and Applications, (pp. 383-419). Florida, CRC Press.
• Turker M., & Arikan, M. (2005). Sequential masking classification of multi-temporal Landsat7 ETM+ images for field-based crop mapping in Karacabey, Turkey. International Journal of Remote Sensing, 26(17), 3813-3830. doi: 10.1080/01431160500166391.
• Utgoff, P. E., & Brodley, C. E. (1990). An incremental method for finding multivariate splits for decision trees. In Machine Learning Proceedings 1990, (pp. 58-65), Morgan Kaufmann.
• Van Niel, T. G., & McVicar, T. R. (2004). Determining temporal windows for crop discrimination with remote sensing: A case study in south-eastern Australia. Computers and Electronics in Agriculture, 45(1-3), 91-108. doi:10.1016/j.compag.2004.06.003.
• Viskovic, L., Kosovic, I. N., & Mastelic, T. (2019, September). Crop classification using multi-spectral and multitemporal satellite imagery with machine learning. In 2019 International Conference on Software, Telecommunications and Computer Networks (SoftCOM) (pp. 1-5). IEEE. doi: 10.23919/SOFTCOM.2019.8903738
• Witharana, C., Civco, D. L., & Meyer, T. H. (2013). Evaluation of pansharpening algorithms in support of earth observation based rapid-mapping workflows. Applied Geography, 37(2013), 63-87. doi: 10.1016/j.apgeog.2012.10.008.
• Yılmaz, V., & Güngör, O. (2013, Mayıs). Görüntü kaynaştırma yöntemlerinde performans analizi. Türkiye Ulusal Fotogrametri ve Uzaktan Algılama Birliği VII. Teknik Sempozyumu, 2013. TUFUAB’2013.
• Zhan, X., Sohlberg, R., Townshend, J., Dimiceli, C., Carroll, M., Eastman, J., Hansen, M. C., & DeFries, R. S. (2002). Detection of land cover changes using MODIS 250 m data. Remote Sensing of Environment, 83(1-2), 336-350. doi: 10.1016/S0034-4257(02)00081-0.
• Zhang, H., & Xu, R. (2018). Exploring the optimal integration levels between SAR and optical data for better urban land cover mapping in the Pearl River Delta. International Journal of Applied Earth Observation and Geoinformation, 64(2018), 87-95. doi: 10.1016/j.jag.2017.08.013.