OPTIMUM SENSOR RESOLUTION FOR PREDICTION OF HOUSING DENSITY 
Bruce Forster 
Director 
Chen Xing and Teck Kim Ng 
Research Students 
Centre for Remote Sensing 
University of New South Wales 
P.O. Box 1, Kensington, NSW 2033 
Australia 
ABSTRACT 
The relationship between sensor resolution and the size of urban features is one of critical importance when 
attempting to use satellite based remote sensing systems for the prediction of urban housing density. Accurate 
prediction of this variable has potential benefits for census and planning studies of urban areas. The visible 
and near infrared reflectance of urban residential areas has been modelled using different housing densities, 
typical of a modern western city. The modelled response of these areas using sensors with resolutions ranging 
from 5 to 30 metres has been determined, and indicates a strong correlation between response variance and 
housing density with an optimum result being achieved at a resolution of 20 metres. These results have been 
applied to selected areas of the Sydney metropolitan area using SPOT XS-mode data. Results and conclusions 
of the study are presented in this paper. 
215