One of the most interesting research topics in the last decade is generating 3 Dimensions (3D) city model,
nevertheless representing a suitable method to achieve the required rapid, automatic, accurate extraction of
realistic model for large urban area used for GIS applications and photogrammetry is still a challenging issue.
Consequently, a new technique and strategy that increase the efficiency for the 3D city modeling is required.
The aim of this research is to develop a simple and efficient semi-automatic approach to generate a 3D city
model for urban area using the fusion of LiDAR data and Ortho-rectified imagery. This integration of these data
sources provides its efficiency for 3D building extraction that represents the main item in the 3D city model.
This approach use both LiDAR data and imagery as the primary cue to delineate building outlines, based on
pixel based classification. The third dimension is obtained automatically from normalized digital surface model
nDSM, and then the 3D model is generated using multi-Faceted patch The accuracy assessment for both height
and building outlines is conducted using the referring to the ground truth. The results of the accuracy assessment
stage illustrated by means of the well-known statistical methods. It is experimentally validated that the
proposed approach can successfully detect different types of buildings from simple rectangle to circular –shape,
when assessed in terms of different quantitative statistics criteria and visual inspection.