Urban LULC Change Detection and Mapping Spatial Variations of Aurangabad City Using IRS LISS-III Temporal Datasets and Supervised Classification Approach

An accurate mapping of urban LULC is essential for urban development and planning. Although urban area represents a little portion of Earth surface, which brings an unbalanced impact on its surrounding areas. However urban LULC mapping and change detection is critical issue by traditional methods. Recent advances in Geospatial technology can be used to map built-up areas for detecting the urban growth patterns. In this work IRS LISS-III sensors image data of 2003, 2009, and 2015 of same season were used. The LULC mapping and change detection was carried out by four supervised classifier namely Maximumlikelihood classifier (MLC), Mahalanobis-Distance (MD), Minimum-Distance-to-Means (MDM), and Parallelepiped classifier (PC). Obtained results were examined by considering the efficiency of each classifier to accurately map the identified LULC classes. It is observed that, MLC has given the highest overall accuracy of 73.07, 83.51, and 93.43% with kappa coefficient of 0.64, 0.78, and 0.90 in 2003, 2009, and 2015 respectively, which are superior among others; hence we have used classified layer obtained from MLC for further change detection and analysis from 2003 to 2015.