Semi-automatic classification for rapid delineation of the geohazard-prone areas using Sentinel-2 satellite imagery

Abstract

The study of land use land cover has become increasingly significant with the availability of remote sensing data. The main objective of this study is to delineate geohazard-prone areas using semi-automatic classification technique and Sentinel-2 satellite imagery in Bhutan. An open-source, semi-automatic classification plugin tools in QGIS software enabled efficient and rapid conduct of land cover classification. Band sets 2-8, 8A, and 11-12 are utilized and the virtual colour composites have been used for the clustering and creation of training samples or regions of interest. An iterative self-organizing data analysis technique is used for clustering and the image is classified by a minimum distance algorithm in the unsupervised classification. The Random Forest (RF) classifier is used for the supervised classification. The unsupervised classification shows an overall accuracy of 85.47% (Kappa coefficient = 0.71) and the RF classifier resulted in an accuracy of 92.62% (Kappa coefficient = 0.86). A comparison of the classification shows a higher overall accuracy of the RF classifier with an improvement of 7.15%. The study highlights 35.59% (512,100 m2) of the study area under the geohazard-prone area. The study also overlaid the major landslide polygons to roughly validate the landslide hazards. Article highlights Semi-automatic classification technique was applied to delineate the geohazard-prone area in the heterogeneous region of Bhutan Himalaya. Unsupervised and supervised classification technique were used to perform land cover classification using the semi-automatic classification plugin (SCP). The Random Forest classifier predicted higher accuracy and the application is rapid and efficient compared to the unsupervised classification.

Understanding the spatio-temporal behaviour of the sunflower crop for subfield areas delineation using Sentinel‐2 NDVI time-series images in an organic farming system

“…The semi-automatic classification plug-in showed high potential and enabled quick image classification using clustering algorithms. Tempa and Aryal [ 53 ] found an overall accuracy of 85.5% of k-means unsupervised heterogeneous on land classification. Furthermore, cluster analysis for the subfield classification was used by many authors on different crops [ [63] , [64] , [65] , [66] , [67] ].…”

“…Furthermore, the Point Sampling Tool was used to extract raster NDVI data values for each georeferenced point. The Semi-automatic classification plugin and the k-means cluster method were used for categorizing the NDVI images [ 53 ]. The Semi-Automatic Classification Plugin with built-in algorithms developed in Python and third-party algorithms for Sentinel-2 pre-processing of images and the post-processing of classifications through ESA SNAP was used [ 51 ].…”

Crop Harvesting Can Affect Habitat Selection of Wild Boar (Sus scrofa)

“…Then, the urbanized area was excluded based on the latest Corine Land Cover layers (https://land.copernicus.eu/ (accessed on 21 November 2021)). Satellite images were classified in Quantum GIS 3.26 using the semi-automatic classification plugin with the Short-Wave Infrared (band combination of SWIR (B12), NIR (B8A), and red (B4)) [28]. Of the distinguished classifications, the "bare ground" class was used, representing the area not covered with vegetation (Supplementary Figure S1).…”

Assessing the Impact of Land Use and Land Cover Change on Air Quality in East Baton Rouge—Louisiana Using Earth Observation Techniques

“…Each tree in the random forest algorithm outputs a prediction based on the class and the class with the highest votes becomes the prediction of the model. By increasing the number of decision trees, the accuracy of the model is also increased [29]. This method can produce a highly accurate classification compared with other commonly used methods [30]…”