Validation of Visually Interpreted Corine Land Cover Classes with Spectral Values of Satellite Images and Machine Learning
We analyzed the Corine Land Cover 2018 (CLC2018) dataset to reveal the correspondence between land cover categories of the CLC and the spectral information of Landsat-8, Sentinel-2 and PlanetScope images. Level 1 categories of the CLC2018 were analyzed in a 25 km × 25 km study area in Hungary. Spect...
| Main Authors: | , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2021-02-01
|
| Series: | Remote Sensing |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2072-4292/13/5/857 |
| _version_ | 1827588315138752512 |
|---|---|
| author | Orsolya Gyöngyi Varga Zoltán Kovács László Bekő Péter Burai Zsuzsanna Csatáriné Szabó Imre Holb Sarawut Ninsawat Szilárd Szabó |
| author_facet | Orsolya Gyöngyi Varga Zoltán Kovács László Bekő Péter Burai Zsuzsanna Csatáriné Szabó Imre Holb Sarawut Ninsawat Szilárd Szabó |
| author_sort | Orsolya Gyöngyi Varga |
| collection | DOAJ |
| description | We analyzed the Corine Land Cover 2018 (CLC2018) dataset to reveal the correspondence between land cover categories of the CLC and the spectral information of Landsat-8, Sentinel-2 and PlanetScope images. Level 1 categories of the CLC2018 were analyzed in a 25 km × 25 km study area in Hungary. Spectral data were summarized by land cover polygons, and the dataset was evaluated with statistical tests. We then performed Linear Discriminant Analysis (LDA) and Random Forest classifications to reveal if CLC L1 level categories were confirmed by spectral values. Wetlands and water bodies were the most likely to be confused with other categories. The least mixture was observed when we applied the median to quantify the pixel variance of CLC polygons. RF outperformed the LDA’s accuracy, and PlanetScope’s data were the most accurate. Analysis of class level accuracies showed that agricultural areas and wetlands had the most issues with misclassification. We proved the representativeness of the results with a repeated randomized test, and only PlanetScope seemed to be ungeneralizable. Results showed that CLC polygons, as basic units of land cover, can ensure 71.1–78.5% OAs for the three satellite sensors; higher geometric resolution resulted in better accuracy. These results justified CLC polygons, in spite of visual interpretation, can hold relevant information about land cover considering the surface reflectance values of satellites. However, using CLC as ground truth data for land cover classifications can be questionable, at least in the L1 nomenclature. |
| first_indexed | 2024-03-09T00:32:42Z |
| format | Article |
| id | doaj.art-9d5a4da099bf463e9ccd50975b9122a4 |
| institution | Directory Open Access Journal |
| issn | 2072-4292 |
| language | English |
| last_indexed | 2024-03-09T00:32:42Z |
| publishDate | 2021-02-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Remote Sensing |
| spelling | doaj.art-9d5a4da099bf463e9ccd50975b9122a42023-12-11T18:24:02ZengMDPI AGRemote Sensing2072-42922021-02-0113585710.3390/rs13050857Validation of Visually Interpreted Corine Land Cover Classes with Spectral Values of Satellite Images and Machine LearningOrsolya Gyöngyi Varga0Zoltán Kovács1László Bekő2Péter Burai3Zsuzsanna Csatáriné Szabó4Imre Holb5Sarawut Ninsawat6Szilárd Szabó7Department of Physical Geography and Geoinformation Systems, Doctoral School of Earth Sciences, University of Debrecen, Egyetem tér 1, 4032 Debrecen, HungaryEnvirosense Hungary Ltd., 4281 Létavértes, HungaryRemote Sensing Centre, University of Debrecen, Böszörményi út 138, 4032 Debrecen, HungaryRemote Sensing Centre, University of Debrecen, Böszörményi út 138, 4032 Debrecen, HungaryDepartment of Physical Geography and Geoinformation Systems, Faculty of Science and Technology, University of Debrecen, Egyetem tér 1, 4032 Debrecen, HungaryInstitute of Horticulture, University of Debrecen, Böszörményi út 138, 4032 Debrecen, HungaryAsian Institute of Technology (AIT), Remote Sensing and Geographic Information Systems (RS&GIS) FoS, Klong Luang, Pathumthani 12120, ThailandDepartment of Physical Geography and Geoinformation Systems, Faculty of Science and Technology, University of Debrecen, Egyetem tér 1, 4032 Debrecen, HungaryWe analyzed the Corine Land Cover 2018 (CLC2018) dataset to reveal the correspondence between land cover categories of the CLC and the spectral information of Landsat-8, Sentinel-2 and PlanetScope images. Level 1 categories of the CLC2018 were analyzed in a 25 km × 25 km study area in Hungary. Spectral data were summarized by land cover polygons, and the dataset was evaluated with statistical tests. We then performed Linear Discriminant Analysis (LDA) and Random Forest classifications to reveal if CLC L1 level categories were confirmed by spectral values. Wetlands and water bodies were the most likely to be confused with other categories. The least mixture was observed when we applied the median to quantify the pixel variance of CLC polygons. RF outperformed the LDA’s accuracy, and PlanetScope’s data were the most accurate. Analysis of class level accuracies showed that agricultural areas and wetlands had the most issues with misclassification. We proved the representativeness of the results with a repeated randomized test, and only PlanetScope seemed to be ungeneralizable. Results showed that CLC polygons, as basic units of land cover, can ensure 71.1–78.5% OAs for the three satellite sensors; higher geometric resolution resulted in better accuracy. These results justified CLC polygons, in spite of visual interpretation, can hold relevant information about land cover considering the surface reflectance values of satellites. However, using CLC as ground truth data for land cover classifications can be questionable, at least in the L1 nomenclature.https://www.mdpi.com/2072-4292/13/5/857LandsatSentinelplanetCLC2018Recursive Feature Eliminationvalidation |
| spellingShingle | Orsolya Gyöngyi Varga Zoltán Kovács László Bekő Péter Burai Zsuzsanna Csatáriné Szabó Imre Holb Sarawut Ninsawat Szilárd Szabó Validation of Visually Interpreted Corine Land Cover Classes with Spectral Values of Satellite Images and Machine Learning Remote Sensing Landsat Sentinel planet CLC2018 Recursive Feature Elimination validation |
| title | Validation of Visually Interpreted Corine Land Cover Classes with Spectral Values of Satellite Images and Machine Learning |
| title_full | Validation of Visually Interpreted Corine Land Cover Classes with Spectral Values of Satellite Images and Machine Learning |
| title_fullStr | Validation of Visually Interpreted Corine Land Cover Classes with Spectral Values of Satellite Images and Machine Learning |
| title_full_unstemmed | Validation of Visually Interpreted Corine Land Cover Classes with Spectral Values of Satellite Images and Machine Learning |
| title_short | Validation of Visually Interpreted Corine Land Cover Classes with Spectral Values of Satellite Images and Machine Learning |
| title_sort | validation of visually interpreted corine land cover classes with spectral values of satellite images and machine learning |
| topic | Landsat Sentinel planet CLC2018 Recursive Feature Elimination validation |
| url | https://www.mdpi.com/2072-4292/13/5/857 |
| work_keys_str_mv | AT orsolyagyongyivarga validationofvisuallyinterpretedcorinelandcoverclasseswithspectralvaluesofsatelliteimagesandmachinelearning AT zoltankovacs validationofvisuallyinterpretedcorinelandcoverclasseswithspectralvaluesofsatelliteimagesandmachinelearning AT laszlobeko validationofvisuallyinterpretedcorinelandcoverclasseswithspectralvaluesofsatelliteimagesandmachinelearning AT peterburai validationofvisuallyinterpretedcorinelandcoverclasseswithspectralvaluesofsatelliteimagesandmachinelearning AT zsuzsannacsatarineszabo validationofvisuallyinterpretedcorinelandcoverclasseswithspectralvaluesofsatelliteimagesandmachinelearning AT imreholb validationofvisuallyinterpretedcorinelandcoverclasseswithspectralvaluesofsatelliteimagesandmachinelearning AT sarawutninsawat validationofvisuallyinterpretedcorinelandcoverclasseswithspectralvaluesofsatelliteimagesandmachinelearning AT szilardszabo validationofvisuallyinterpretedcorinelandcoverclasseswithspectralvaluesofsatelliteimagesandmachinelearning |