Date of Award
Master of Science
Laura F. Serpa
The use of 3-D Structure from Motion models to study fossils is a relatively new technique that has an advantage over traditional field methods, such as plaster casts or collecting samples, because it is non-destructive and can be easily shared with a wide audience. It is also ideal for use with persons with disabilities because it does not require field access and it can be manipulated for optimum computer viewing or printed with a 3-D printer for tactile examination as needed. Our study focused on a known site in Cristo Rey, New Mexico where Cretaceous Iguanadon and Ankylosaur footprints have been studied extensively for the past 20 years but exposure to the elements and tourists were producing significant erosion of the fossils. We collected nearly 10,000 high-resolution photographs in an 8m x 9m area with a high-resolution hand held camera. This proved to be too large of a data set for our computer facilities to process so we downsized the dataset to approximately 2500 photos that covered approximately one fourth of the target area and built a point cloud from those photos. Even at a relatively low quality setting it took approximately 70 hours of computer time to build the model. The model was rendered into a 3D PDF, a DEM, and a number of other formats for additional manipulation.
In addition to the hand held photos, we used a Go Pro camera mounted on a UAV to capture approximately 100 images of the same area for comparison with the high-resolution data. Both data sets were processed in a similar manner and each brought out some important feature of the site. We were surprised when our examination of the models showed several footprints and
tracks that had not previously been identified. We believe that further work
with the models will lead to the identified more new life forms preserved in the shallow muds that made up the Cretaceous period in this area.
Received from ProQuest
Valeria V. Martinez
Martinez, Valeria V., "Reconstructing Dinosaur Foot Tracks and Identifying New Footprints using Structure from Motion Photogrammetry" (2016). Open Access Theses & Dissertations. 689.