STRUCTURE FROM MOTION-MULTIVIEW STEREO (SFM-MVS) PHOTOGRAMMETRY FOR CONSTRUCTING VIRTUAL OUTCROP MODELS AND CLASSROOM “FIELD” EXPERIENCES

We present an educational workflow focused on learning structural geology concepts which until recently would have been relegated only to the field. The exercise revolves around a point cloud created via a SfM-MVS workflow as well as a geologic model derived from the point cloud. Modern computing power has made the creation and manipulation of dense point clouds (10^6 points or greater) routinely available for universities and their students. This advancement of technology has significant implications for the use of virtual outcrop models (VOMs) in the classroom. Among the most useful is the ability to quickly create VOMs of an educational outcrop or scene which can then be investigated by students at the home institution.

In the case-study presented here the outcrop imaged is in the Ibex Hills, CA and exhibits two prominent fold generations resulting in refolded folds. Students are initially instructed to analyze individual folds from traditional field photographs before being introduced to the 3D model to address shortcomings in their initial 2D focused hypotheses. Bedding orientations derived from the VOM were also provided for students to further interpret the geology. A texture derived from the point cloud was hosted in a browser based platform which can be used to manipulate the VOM to provide context for further interpretation. Ultimately, the goal was to display the point cloud, or an associated texture, within Google Earth to be viewed and interpreted alongside regional structures. However, the size of the point cloud made this unmanageable and instead a simplified model of the folds was inserted as a KMZ file into Google Earth.

In its current state the exercise presented here highlights the advantages of 3D VOMs in teaching complex structural outcrops and the role that SfM photogrammetry plays in that sphere. However, there are limitations on what can be used to interface with students due to the size of dense point clouds. Possible avenues include browser based platforms to host texturized point clouds or open source point cloud software. While there are advantages to using point cloud processing software, namely being able to directly interact with the data, it does require more time for educators and teachers to adapt to and there may be issues of sufficient computing power for classroom settings.