Abstract:
Automatic change detection and open space area
extraction in urban environment is one of the crucial
components towards the efficient updating of
Geographic Information System (GIS), government
decision-making, urban land management and
planning. Original Morphological Building Index
(MBI) can extract interest building features for multitemporal
high-resolution satellite image but this
approach wrongly classified as buildings. In this
paper, jointly approach of modified MBI, Normalized
Different Vegetation Index (NDVI) and Entropy is
developed for identifying low quality satellite images
over different years. Then, matching-based change
rule is applied to obtain changes area of urban region.
The proposed method is insensitive to the geometrical
differences of buildings caused by different imaging
conditions and is able to significantly reduce false
alarms and also achieves much improved detection
accuracy and overall performance. The effectiveness
of the method is validated by comparing with MBIbased
Change Vector Analysis (CVA) and
Multivariate alteration detection (MAD) transformation.
