Utilization of Advanced Slope-based Indexing Technique (ASIT) for the extraction of built-up land

Abstract

ABSTRACTCurrently, there are numerous remote-sensing indices that are used in the classification of built-up land in order to monitor population and urbanization growth. As urbanization is the cause of Urban Heat Island (UHI) as well as a major contributor to climate warming, therefore, automatic mapping of built-up areas is considered an important tool to monitor the effects of urbanization on the climate change. The current research study proposed a new built-up indexing technique, i.e. Advanced Slope-based Indexing Technique (ASIT) for separating land features having similar spectral pattern and explores its potential for mapping built-up area in comparison to other available spectral indices for the study area comprises of Karachi region of Pakistan. The Otsu thresholding method is utilized on ASIT built-up index as well as on four of the available built-up indices for the separation of built-up areas from non-built-up areas. To analyse the success rate of the ASIT built-up index, 850 reference points are taken from high-resolution pan-chromatic satellite data. The implementation of the proposed index to a portion of Landsat 8 image consisting of study area results in an overall accuracy (OA) of 89.566% with a Kappa coefficient (K) of 0.779. The comparative analysis inferred that the proposed ASIT built-up index improves the OA of 2.931% and K of 0.053 when compared with the VbSWIR1-BI index which is the supposed to be the most effective index among all the available indices.