A novel approach of efficient 3D reconstruction for real scene using unmanned aerial vehicle oblique photogrammetry with five cameras

Abstract

In recent years, unmanned aerial vehicle (UAV) oblique photogrammetry is a new and advanced technology in the field of surveying and mapping. In this paper, an oblique photogrammeter with five high-precise measuring cameras is developed, which can be controlled by a microcontroller, positioned by Global Positioning System (GPS) and BeiDou Navigation Satellite System (BDS), and accessed through narrowband internet of things (NB-IoT) module interface. Meanwhile, distributed parallel computing and graphics processing unit (GPU) are used to improve the efficiency of 3D modeling for the real scene. Finally, the urban spatial information collected by UAV oblique photogrammetry and the real scene 3D modeling results are verified by experiments. The experiment results demonstrate that the method proposed in this paper can efficiently complete the collection of urban spatial information and improve the efficiency 3D modeling for the real scene. The proposed method has about 30% improved efficiency compared to the conventional methods.