“…As with conventional GPS‐supported aerial triangulation, if UAV imagery comprises a strong block geometry, with good, consistent scene texture to enable a dense and regular distribution of image tie points (for example, obtained using the scale‐invariant feature transform (SIFT) algorithm of Lowe, ), camera attitudes may be computed in the bundle block adjustment with high accuracy. Moreover, in this configuration, the relatively low accuracy of current lightweight IMU observations offer little additional constraint in the bundle block adjustment, as per conventional integrated sensor orientation (Rehak and Skaloud, ; Stöcker et al., ). For these reasons, approaches to remove GCPs in UAV photogrammetry are usually concerned with camera positioning only (Gerke and Przybilla, ), and such GPS‐supported aerial triangulation will be referred to in this paper hereafter as direct georeferencing.…”