Mean and standard deviation of the point cloud to

In the next step, advantages of different geometric configurations from both data sets should be utilised to fill potential data gaps to receive AGN 194310 consistent digital 3D model of the gully. Therefore, data fusion was performed with UAV data and fine registered TERRA point clouds (Fig. 7). Due to the geometric configuration of the image acquisition from the UAV platform (nadir perspective), the morphology of the steep sidewalls could only be mapped fragmentary and less detailed in the UAV model. In contrast, for the terrestrial data image matching in the gully bed, as well as in the flat terrain, was prone to error due to a very oblique perspective with a very low incidence angle. However, for the steep gully walls the 3D reconstruction operated very well. Hence, with the TERRA model gaps in the UAV model were closed. Furthermore, the micro relief is cambium clearly visible due to the higher ground sample distance. Positive deviations (red — small convexities), as well as negative deviations (dark green and blue — undercuts), could be reconstructed by the terrestrial survey and transferred to the homogenous and consistent model by data fusion (Fig. 7).