B41D-0464
UAS-Borne Photogrammetry for Surface Topographic Characterization: A Ground-Truth Baseline for Future Change Detection and Refinement of Scaled Remotely-Sensed Datasets

Thursday, 17 December 2015
Poster Hall (Moscone South)
Emily S Schultz-Fellenz1, Ryan Coppersmith2, Aviva J Sussman3, Steve Vigil4, Robert Dzur5, Katherine Norskog1, Richard Kelley1 and Liz Miller1, (1)Los Alamos National Laboratory, Los Alamos, NM, United States, (2)Coppersmith Consulting Inc., Walnut Creek, CA, United States, (3)Los Alamos National Laboratory, Geophysics Group, Los Alamos, NM, United States, (4)Sandia National Laboratories, Albuquerque, NM, United States, (5)Bohannon Huston, Inc., Albuquerque, NM, United States
Abstract:
While long-term objectives of monitoring and verification regimes include remote characterization and discrimination of surficial geologic and topographic features at sites of interest, ground truth data is required to advance development of remote sensing techniques. Increasingly, it is desirable for these ground-based or ground-proximal characterization methodologies to be as nimble, efficient, non-invasive, and non-destructive as their higher-altitude airborne counterparts while ideally providing superior resolution. For this study, the area of interest is an alluvial site at the Nevada National Security Site intended for use in the Source Physics Experiment’s (Snelson et al., 2013) second phase. Ground-truth surface topographic characterization was performed using a DJI Inspire 1 unmanned aerial system (UAS), at very low altitude (< 5-30m AGL). 2D photographs captured by the standard UAS camera payload were imported into Agisoft Photoscan to create three-dimensional point clouds. Within the area of interest, careful installation of surveyed ground control fiducial markers supplied necessary targets for field collection, and information for model georectification. The resulting model includes a Digital Elevation Model derived from 2D imagery. It is anticipated that this flexible and versatile characterization process will provide point cloud data resolution equivalent to a purely ground-based LiDAR scanning deployment (e.g., 1-2cm horizontal and vertical resolution; e.g., Sussman et al., 2012; Schultz-Fellenz et al., 2013). In addition to drastically increasing time efficiency in the field, the UAS method also allows for more complete coverage of the study area when compared to ground-based LiDAR. Comparison and integration of these data with conventionally-acquired airborne LiDAR data from a higher-altitude (~ 450m) platform will aid significantly in the refinement of technologies and detection capabilities of remote optical systems to identify and detect surface geologic and topographic signatures of interest. This work includes a preliminary comparison of surface signatures detected from varying standoff distances to assess current sensor performance and benefits.