CLASSIFICATION OF AURANGABAD CITY USING HIGH RESOLUTION REMOTE SENSING DATA

Kiran Bagade, Second Amol Vibhute, K.V. Kale

Abstract: The current study highlights the advantages of remote sensing and Geographic Information System (GIS) in the field urban planning and management. IRS-P6 Resourcesat-1 LISS-IV high spatial resolution (5.8m) data with three spectral bands were used for urban classification. The study area Aurangabad is the capital metro city of Maharashtra State, India. ENVI 4.4 image processing tool was used for classification of satellite data on the basis of supervised approach. Two statistical algorithms were used for urban classification such as Minimum distance and Mahalanobis distance classifier. Lastly the accuracy of the classification was performed through ground truth. The result indicates that the Minimum distance classifier gives the better results than Mahalanobis classifier which are 80.2817% and 70.4225% respectively. Hence it is identified minimum distance is best for urban classification

Keywords: Supervised classification, Multispectral, Confusion matrix, Producer’s accuracy, Users accuracy.

DOI: https://doi.org/10.15623/ijret.2015.0401052