Photogrammetry is best defined as a reversal of the photographic process that captures the 3D world in two dimensions. By taking thousands of regimented, overlapping photos of my ruined sites and processing them with the open souce software VisualSFM, I was able to capture dense 3D data of the ruined foundations of two of my sites. I captured this data in March 2013 using a handheld DSLR, sometimes mounted on a 15 ft. pole, as well as a small point and click camera attached to the line of a kite that I flew over the site.
Above: This video is an export from the CMPMVS software that meshes and textures a dense point cloud after it has been processed in VisualSFM. This video simultaneously shows the original photo at top left, the dense point cloud, un-textured surface mesh, and textured mesh at the bottom right.
Merton Spire, carved for Oxford’s Merton College Chapel in 1451. Given to the University of Virginia In 1928. Photogrammetry experiment processed using Agisoft “Photoscan.”
Varsity Hall at UVa. Built in 1858 as an infirmary. Photogrammetry experiment processed in Autodesk’s “123D Catch.”
Westover Estate. Circa 1915 Greek revival house now owned by the University of Virginia. Photogrammetry experiment processed in Autodesk’s “123D Catch.”
Dense surface model of Lord Botetourt statue at the College of William and Mary. Photogrammetry experiment processed in Agisoft “Photoscan.”
Point cloud of the fortress-monastery of Montesa in Valencia. This cloud was processed from more than 10 thousand images using ChangChang Wu’s VisualSFM and registered/compiled using Cloud Compare. The cloud is composed of more than 91 million points.
A Photogrammetric point cloud of Calatrava la Nueva. This cloud was processed from more than 16 thousand photos using ChangChang Wu’s VisualSFM software and registered using Cloud Compare. The cloud is composed of more than 205 million points.