Evaluate appropriateness of Sentinel-2 optical data for extracting CORINE land cover classes
Abstract
Nowadays, every month/year changes occur on the surface of the Earth from a small build ups till large-scale natural disasters that destroyed entire settlements. Observation and analysis using the satellite data can help to improve and prevent any natural or anthropogenic disaster, plan and calculate beneficial usage of the land cover. Moreover, there are huge amount of satellites which provides special data for monitoring the Earth. One of them, Sentinel-2 is latest one of the sensors which provides high resolution data (10m, 20m, 60m) with 13 bands total [Sentinel-2]. The purpose of this paper is to evaluate the role of Object-based Image Analysis (OBIA) to extract CORINE Land Cover Classes (CLC) from Sentinel-2 data by comparing the classification results yielded by three widely used classification methods, namely Support Vector Machine (SVM), Random Forest (RF) and K-Nearest Neighbour (KNN). Since, Sentinel-2 is the newest optical data provider, it is one of the first approaches to use Sentinel-2 for extracting CLC classes.
Keywords
Remote sensing, accuracy assessment, supervised classification, eCognition, machine learning algorithm
Figure 1 below illustrates final poster for the project.
Abstract
Nowadays, every month/year changes occur on the surface of the Earth from a small build ups till large-scale natural disasters that destroyed entire settlements. Observation and analysis using the satellite data can help to improve and prevent any natural or anthropogenic disaster, plan and calculate beneficial usage of the land cover. Moreover, there are huge amount of satellites which provides special data for monitoring the Earth. One of them, Sentinel-2 is latest one of the sensors which provides high resolution data (10m, 20m, 60m) with 13 bands total [Sentinel-2]. The purpose of this paper is to evaluate the role of Object-based Image Analysis (OBIA) to extract CORINE Land Cover Classes (CLC) from Sentinel-2 data by comparing the classification results yielded by three widely used classification methods, namely Support Vector Machine (SVM), Random Forest (RF) and K-Nearest Neighbour (KNN). Since, Sentinel-2 is the newest optical data provider, it is one of the first approaches to use Sentinel-2 for extracting CLC classes.
Keywords
Remote sensing, accuracy assessment, supervised classification, eCognition, machine learning algorithm
Figure 1 below illustrates final poster for the project.