3D building Reconstruction from Google Earth Satellite Images

Abstract

The system is aimed to develop the 3D building model of urban area with low resolution satellite images from Google earth that provides to create the three dimensional geographic information systems (GIS). In this paper, an efficient building extraction algorithm is introduced based on color segmentation and morphological method. The approach is based on the affine projection to compute the corresponding between two satellite images. To reconstruct the 3D model, the depth map is created by the modified factorization algorithm with singular value decomposition (SVD) is used to project the image point. The objective of the system is to reconstruct the 3D model of building in urban area with low resolution satellite images using efficient digital image processing method.