DOI: 10.36347/sjet.2019.v07i05.004

Pages: 180–188

There is an increasing interest in generating virtual reality invironments due to the growing 3D visual media in the market. Aerial photogrammetric data is usually used to generate virtual urban models but there are some issues in creating 3D city model from aerial photos. This includes problems in identifying small buildings and limited building geometry detail levels (LOD) especially for building facades. To progress aerial photogrammetric data for the development of virtual city modeling, this research is carried out by integrating multi-source image-based data to verify LOD3 level of buildings accomplished by Close Range Photogrammetry (CRP). This study includes the integration of aerial and CRP images those delivered from Unmanned Aerial Vehicle (UAV) platform and hand-held static camera, for the construction of augmented reality model of the University of Baghdad (UOB) campus. Two buildings with different specifications inside the university campus are selected to apply the study and discuss results. In this study, the integration of multi-source image-based data, are achievable using different algorithms from commercial software. The results approved that using UAV platform to capture close-range images of very high building facades is recommended and can deliver superior results as compared with results delivered from static ground stations using hand-held cameras. The validation accuracy of the final virtual reality UOB model deliver RMSE of 0.271051m., 0.250703 m. and 0.301240 m. in easting, northing, and elevation respectively, which consider promising for small scale projects with small detailed objects in urban areas.