Published Paper Details:

Vegetation extraction from remote sensing imagery is the process of extracting vegetation information by interpreting satellite images based on the interpretation elements such as the image colour, texture, tone, pattern and association information. The vegetation maps are critical for understanding biodiversity management and planning from local to global scales. There are two essential elements of vegetation maps that are based on classification of vegetation and spatial attribution of that classification. The difference between these vegetation cover can be done with using remote sensing and GIS. Remote sensing classification is an essential technique for mapping forest vegetation attributes like floristic composition, biomass, tree density etc. that addresses the considerable logistical challenge of measuring, mapping and monitoring across very large areas in an accurate, repeatable and cost-effective way. The most basic method is identifying plant communities or species is to create a vegetation map. Machine learning (ML) is a subdivision of artificial intelligence in which the machine learns from machine-readable data and information. It uses data, learns the pattern and predicts the new outcomes. Its popularity is growing because it helps to understand the trend and provides a solution that can be either a model or a product. This study is carried out to perform various machine learning algorithms for vegetation classification using remote sensing and ancillary spatial data across a heterogenous forest ecosystem in Virajpet Taluk, Kodagu district. The recently developed sentinel-2 satellite imagery holds great potential for improving the vegetation type classification at medium to large scales due to the concurrent availability of multispectral bands with high spatial resolution and quick revisit time. A study is classifying using Supervised classification, a type of machine learning technique. To perform the supervised classification, training samples have been generated using ARCGIS to run the spatial model in ERDAS software. The machine learning algorithm such as support vector machine (SVM), Random Forest (RF), k-Nearest Neighbour (k-NN), Navies Bayes and CART were used to classify. In terms of classification accuracy, a combination of these parameters was tested: The number of trees (NT), the variables per split (VPS) and the Bag Fraction (BF). Thus, this studies key contribution is classification of vegetation types and how the ML techniques are an efficient approach to map different land use and land cover classes, including different vegetation types and demonstrated that some algorithms perform this task with higher accuracy than the others. In our investigation, the RF algorithm chieved satisfactory results when compared with others and characterized with more accurate rate and less error rate.