CORINE Level-2 Land Cover Classification from Sentinel-2 Images Using U-Net Deep Learning Model

Year 2025, Volume: 12 Issue: 2, 34 - 44, 30.06.2025

Emin Atabey Peker , Murat Çalışkan , Eser Bora , Çiğdem İnan Mehmet Erkan Uçaner

Abstract

This study investigates the automated classification of CORINE Level 2 land cover classes using deep learning-based semantic segmentation in Türkiye. We implemented an advanced U-Net architecture incorporating spatial attention mechanisms and residual connections, utilizing multi-spectral Sentinel-2 imagery (RGB, NIR, and SWIR bands). A total of 92 working units were identified across Türkiye to represent all 15 classes of Level 2, covering 12,000 square kilometers (approximately 1.5% of the country's total area). The imagery was collected between June and September 2018, corresponding to the original CORINE 2018 production timeframe. The model was trained on 2,815 patches of 256x256 pixels generated from these working units. Our architecture integrates an Atrous Spatial Pyramid Pooling (ASPP) block for multi-scale feature extraction and employs a hybrid loss function combining Categorical Cross-Entropy and Jaccard Loss. The model achieved a mean IoU of 0.5018 across all classes, with particularly strong performance in water bodies (Marine Waters: 0.9077 IoU, Inland Waters: 0.8323 IoU) and permanent crops (0.6855 IoU). While the model excelled in classifying spectrally distinct classes, it faced challenges with classes sharing similar characteristics, such as artificial non-agricultural vegetated areas (0.1273 IoU) and pastures (0.1760 IoU). This research demonstrates the potential of deep learning approaches for automated CORINE land cover mapping, while highlighting areas for future improvement in distinguishing between spectrally similar classes.

Keywords

CORINE , Deep Learning , U-Net , Remote Sensing , Landcover

References

• Balado, J., Arias, P., Vilariño, L. D., & González-deSantos, L. M. (2018). Automatic CORINE land cover classification from airborne LIDAR data. Procedia Computer Science, 186-194.
• Balzter, H., Cole, B., Thiel, C., & Schmullius, C. (2015). Mapping CORINE Land Cover from Sentinel-1A SAR and SRTM Digital Elevation Model Data using Random Forests. Remote Sensing, 14876-14898.
• Copernicus Land Monitoring Service (CLMS). (2025). Corine land cover. Retrieved from https://land.copernicus.eu/en/products/corine-land-cover
• Demir, D. B., & Musaoglu, N. (2023). Automatic Classification Of Selected Corine Classes Using Deep Learning Based Semantic Segmentation. In The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences (Volume XLVIII-M-3-2023). Denver, Colorado: ASPRS.
• Gorelick, N., Hancher, M., Dixon, M., Ilyushchenko, S., Thau, D., & Moore, R. (2017). Google Earth Engine: Planetary-scale geospatial analysis for everyone. Remote Sensing of Environment, 202(3), 18-27.
• Johnson, K. E., & Koperski, K. (2017). WorldView-3 SWIR Land Use-Land Cover Mineral Classification: Cuprite, Nevada. In Pecora 20 - Observing a Changing Earth; Science for Decisions--- Monitoring, Assessment, and Projection. Sioux Falls, SD.
• Loshchilov, I., & Hutter, F. (2017). Decoupled Weight Decay Regularization. In Proceedings of the International Conference on Learning Representations. https://doi.org/10.48550/arXiv.1711.05101
• Macarringue, L. S., Bolfe, É. L., & Pereira, P. R. (2022). Developments in Land Use and Land Cover Classification Techniques in Remote Sensing: A Review. Journal of Geographic Information System, 14, 1-28.
• Özturk, O., Sarıtürk, B., & Seker, D. Z. (2020). Comparison of fully convolutional networks (FCN) and U-Net for road segmentation from high resolution imageries. International journal of environment and Geoinformatics, 7(3), 272-279.
• Ronneberger, O., Fischer, P., & Brox, T. (2015). U-Net: Convolutional Networks for Biomedical Image Segmentation. In N. Navab, J. Hornegger, W. Wells, & A. Frangi (Eds.), Medical Image Computing and Computer-Assisted Intervention (pp. 234-241). Springer.
• Solórzano, J. V., Mas, J. F., Gao, Y., & Cruz, J. G. (2021). Land Use Land Cover Classification with U-Net: Advantages of Combining Sentinel-1 and Sentinel-2 Imagery. Remote Sensing, 13(18), 3600.
• Stara, L., & Halounova, L. (2022). Classification Of Selected Corine Classes Using Sentinel-2 Data. In The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences (Volume XLIII-B3). Nice, France: XXIV ISPRS Congress.
• U.S. Environmental Protection Agency (EPA). (2024). Report on the Environment (ROE): What are the trends in land cover and their effects on human health and the environment?
• Varga, O. G., Kovács, Z., Bekő, L., Burai, P., Szabó, Z. C., Holb, I., . . . Szabó, S. (2021). Validation of Visually Interpreted Corine Land Cover Classes with Spectral Values of Satellite Images and Machine Learning. Remote Sensing, 13(5), 857.
• Vinay, S. B. (2024). AI and Machine Learning Integration with Aws Sagemaker: Current Trends and Future Prospects. International Journal of Artificial Intelligence Tools (IJAIT), 1-24.
• Zhang, C., & Li, X. (2022). Land Use and Land Cover Mapping in the Era of Big Data. Land, 11, 1692.
• Zhao, S., Tu, K., Ye, S., Tang, H., Hu, Y., & Xie, C. (2023). Land Use and Land Cover Classification Meets Deep Learning: A Review. Sensors, 23(21), 8966.