Structure-from-Motion is now a standard technique for documenting outcrops and paleoseismic trenches. It is also widely used in archaeology, architecture, disaster response, etc. In a perfect world, we would always have lots of well-known ground control points (GCPs) for our 3D models, for example by using a differential GPS. However, quite often we can’t do that because we don’t have a dGPS, because we stumbled upon an outcrop by accident, because we don’t want to carry the tripod to the top of a summit, etc. The problem is that without GCPs we might end up with funny errors in our model. Ian Pierce has written a tutorial about how to overcome many of these problems by combining SfM with the scans of the iPad’s and iPhone’s built-in lidar sensor: https://neotectonic.github.io/posts/iOS_Surveying/
Thanks Ian!!