Land use / Land Cover (LULC) is a significant factor which plays a vital role in defining an urban ecosystem. Interpretations of LULC are eased in recent times by utilizing hyperspectral and multispectral datasets obtained from airborne and spaceborne platforms. In this study, an attempt has been made to comparatively assess the potentiality of AVIRIS NG hyperspectral data with Sentinel 2 multispectral data through applied classification techniques for Kalaburagi urban sphere. Hyperspectral data being acquired airborne consists of 425 bands covering wavelength from 356 nm to 2500 nm are analyzed for dimensionality reduction transform and are further classified. Meanwhile Sentinel 2 multispectral dataset being spaceborne and less expensive having a minimal ground sampling distance of 10 meter are classified. Spectral responses of both multispectral and hyperspectral bands were analyzed to derive reflectance spectra for considered spectral bands that are well – distributed among all the other spectral bands. The most relevant information which significantly constitutes urban land cover is considered thus avoiding redundancy among the bands. This paper focuses on applying standard supervised classification algorithms associated with dimensionality reduction techniques. Spectral resampling of hyperspectral to multispectral data product have been used to assign user – defined classes. For performance evaluation, the results are validated in order to check which of the given datasets outperforms well and provides better classified results.
Previous Article in event
Integrated hydrological modelling over upstream catchments of Himalayan rivers and assessment of extreme hydrological eventsPrevious Article in session
Next Article in event
Assessment on the potential of multispectral and hyperspectral datasets for Land Use / Land Cover classification
Published: 05 June 2019 by MDPI in 2nd International Electronic Conference on Geosciences session Earth Sciences through Earth Observation
Keywords: AVIRIS NG; Sentinel 2; dimensionality reduction; spectral resampling; Kalaburagi