Urban land use dynamics and change implication by using low-resolution LISS III imagery

Abstract

Extension of urban areas is augmented reality. Identification of urban areas and mapping urban sprawl is directly related to population growth. Very frequent identification and monitoring of urban change are needed for urban governance and planners. The objective of the proposed paper was to extract the urban land cover by locating the human settlement areas and to develop a model to map potential spatio-temporal urban expansion. Multi-temporal imagery of Kalaburgai, state of Karnataka, India has been used for the work. The imagery was subjected for supervised fuzzy classification to extract spatial information of the urban area, Principal Component Analysis to transform the correlated components and lastly, unsupervised post classification on composite PCA images to separate uncorrelated pixels. Uncorrelated pixels depict the temporal change data. Results were satisfying, with 5.53 percentage area of urban change implications. Finally, the accuracy assessment of the proposed method shows promising results with an overall accuracy to be 80% and 0.78 for the kappa coefficient. Quantification of urban areas and probabilities of villages changing into cities can be done accurately by the proposed method. Significant urban expansion was observed by the results in Kalaburagi city from 2010 to 2013 with mostly vegetation changing into urban encroachment.