Dinesh Sathyamoorthy
Research Officer
Science & Technology Research Institute for Defence
Malaysia

Biography
Dinesh Sathyamoorthy received the B.Eng. and M.Eng.Sc. degrees in computer engineering from Multimedia University (MMU) in 2003 and 2006 respectively, and the PhD degree in electrical & electronics engineering from Universiti Teknologi Petronas (UTP) in 2013. He is currently a research officer in the Science & Technology Research Institute of Defence (STRIDE), Ministry of Defence. He serves on the editorial boards for the Defence S&T Technical Bulletin, Malaysian Journal of Remote Sensing & GIS, and The Journal of Defence and Security, as well as reviewer for several international journals and conferences, including the Defence Science Journal, Engineering Applications of Artificial Intelligence, International Journal of Water Resources and Environmental Engineering, and the annual IEEE International Geoscience & Remote Sensing Symposium (IGARSS). He is also a committee member of the Institution of Geospatial & Remote Sensing Malaysia (IGRSM) and IEEE Geoscience & Remote Sensing (GRSS) Malaysia Chapter. His research interests include geospatial technologies and digital signal & image processing.

Abstract
Comparative Analysis of Cell- and Object-Based Local Variance for Physiographic Features Extracted from Multiscale Digital Elevation Models

While a number of methods have been proposed to study the effect of scale on intrinsic spatial attributes of terrains, from the aspect of simplicity and performance, local variance is more widely used for geospatial analysis. This paper is aimed at providing a comparative analysis of cell- and object-based local variance LV for physiographic features extracted from multiscale digital elevation models (DEMs). The results obtained highlight the advantages of object-based LV over cell-based LV. LV computation using the cell-based approach does not take into account the spatial and contextual information of the cells analysed, resulting in assignment of LV values that do not quantify the change in spatial variabilities of the multiscale DEMs and corresponding physiographic features. By analysing objects rather than individual cells, the object-based approach takes into account the spatial and contextual information of the cells, resulting in LV values that accurately quantify the change in spatial variabilities of the physiographic features over the scales.