CLASSIFICATION OF CROPS ON SENTINEL-2 IMAGES BY MACHINE LEARNING METHODS
DOI:
https://doi.org/10.24867/02RB02PejakKeywords:
Classification, Machine Learning, Sentinel-2, Random Forest, Vegetation IndicesAbstract
The thesis evaluates crops classification methods using Sentinel-2 satellite images with twelve vegetation indexes as additional features. Data includes time-series of multispectral images of AP Vojvodina in the period from March to September 2016, taken from the two corresponding satellite paths, R036 and R136. For the classifier training purposes, the information was collected on the type of crops planted on a number of fields. The vegetation indices contributed to the improvement of the classifier performance of 1 % in all experiments, achieving the overall accuracy of 95 %.
References
[1] Lugonja P., Marko O., Panić M., Brkljač B., Brdar S., Crnojević V.: Sentinel-2 and Landsat-8 for highresolution land cover mapping in sustainable agriculture, 8. WorldCover Conference, European Space Agency (ESA), GEO, FAO, and EU, Rim: European Space Agency (ESA), 14-16 Mart, 2017
[2] Brkljač B., Lugonja P., Minić V., Brdar S., Crnojević V.: Data enrichment of Sentinel-2 and Landsat-8 surface-reflectance measurements for agriculture oriented services, 3. Earth Observation Open Science Conference, European Space Agency (ESA), Rim: European Space Agency (ESA), 25-28 Septembar, 2017
[3] Sentinel-2 User Handbook, https://sentinel.esa.int/documents/247904/685211/Sentinel-2_User_Handbook
[4] Xue, J. and Su, B., 2017. Significant remote sensing vegetation indices: A review of developments and applications. Journal of Sensors, 2017.
[5] Breiman, L., 2001. Random forests. Machine learning, 45(1), pp.5-32.
[6] James, G., Witten, D., Hastie, T. and Tibshirani, R., 2013. An introduction to statistical learning (Vol. 112). New York: springer.
[2] Brkljač B., Lugonja P., Minić V., Brdar S., Crnojević V.: Data enrichment of Sentinel-2 and Landsat-8 surface-reflectance measurements for agriculture oriented services, 3. Earth Observation Open Science Conference, European Space Agency (ESA), Rim: European Space Agency (ESA), 25-28 Septembar, 2017
[3] Sentinel-2 User Handbook, https://sentinel.esa.int/documents/247904/685211/Sentinel-2_User_Handbook
[4] Xue, J. and Su, B., 2017. Significant remote sensing vegetation indices: A review of developments and applications. Journal of Sensors, 2017.
[5] Breiman, L., 2001. Random forests. Machine learning, 45(1), pp.5-32.
[6] James, G., Witten, D., Hastie, T. and Tibshirani, R., 2013. An introduction to statistical learning (Vol. 112). New York: springer.
Downloads
Published
2019-05-02
Issue
Section
Biomedical Engineering
