摘要: UAVs (Unmanned Aerial Vehicle systems) are mostly low cost systems and flexible and therefore a suitable alternative solution for aerial photogrammetry at a large scale compared with other mobile mapping systems. Compared to metri... 展开 UAVs (Unmanned Aerial Vehicle systems) are mostly low cost systems and flexible and therefore a suitable alternative solution for aerial photogrammetry at a large scale compared with other mobile mapping systems. Compared to metric camera systems, non-metric digital cameras are relatively inexpensive, readily avalailable, small volume, light weigh. Therefore,they are suitble for application in mini UAVs. It is a foreland task to integrate both and form a new type photogrammetric system. The contribution of this paper is fourfold. First, we will introduce the system of "UAVRS-Ⅱ". Second, we will address the camera calibration and distortion analysis process for the non-metric cameras using in the system. Third, we proposes the automatic aerial triangulation process. In this part, we put forward the method to reject rough error in aerial triangulation step by step based on trifocal tensor in computer vision as well as the free net bundle adjustment in photogrammetry. Fourth, we not only generate the digital orthophotos at scale of 2000 with aerial photographs collected from a test flight executed in the city of Huishui ,Guizhou province at average scale of 1:9 000, but also give the accuracy of the results of aerial triangulation and the orthophotos. 收起