HIGH RISE BUILDING IDENTIFICATION FROM SPOT 6 MULTISPECTRAL AND DIGITAL SURFACE MODEL (DSM) USING OBJECT BASED IMAGE ANALYSIS

Zylshal, Zylshal
Nugroho, Jalu Tejo
Prasasti, Indah

Jurnal Geografi : Media Informasi Pengembangan dan Profesi Kegeografian
Vol 14, No 2 (2017): July 2017

This study focuses on one aspect of urban geometry called urban canyon. Urban canyon defined by a relatively narrow street lined by tall buildings. The initial step to extract the urban canyon is to identify the tall buildings. This study aims to discuss the potential use of the SPOT-6 multispectral data and its digital surface model (DSM), using object-based image analysis methods and terrain analysis, to identify the high-rise buildings in some part of Jakarta, Indonesia. Using slope and elevation percentile from the DSM as well as the spectral information of the SPOT-6 image, we then processed using the Object Image Analysis (OBIA) method and decision tree algorithm (crisp classification), we are able to obtained the identification rate of 78% with mean location accuracy of 30 meter (5 pixels).