Abstract:
Rapid urbanization and urban growth is
continuing to be one of the crucial issues of global
change affecting physical dimensions of cities. In this
study, building growth change detection is investigated
as buildings are one of the most dynamic structures in
urban areas. The modified MBI is applied to extract
building features to know how much area has
changed.In this system, height information is not
considered for building extraction because of without
using multispectral band images and Depth.To
accurate the position of building extraction index in
change detection process, image registration is used
that seeks to remove the two-date images geometric
inconsistent object using SURF and RANSAC. It is
significantly reduce error rates and improve overall
accuracy of change detection process.The experiments
show that the proposed change detection algorithms
can achieves satisfactory correctness rates with a low
level of error rate, and give better result than SIFT
feature extraction method in image registration.
