Photogrammetry, in short, is the practice of making measurements from photographs. Taking digital photographs in a planned sequence, then having those photographs digested by specialty software which stitches the photos into a model is the most common technique. The output is typically a drawing, map, or some graphical construct where one can query and obtain useful data. The use of photogrammetry in documenting property loss situations is not necessarily new, but recent developments in software leading to richer and more useful “models” are noteworthy for our industry.
During the documentation of a claim through photogrammetry, the end result is a viewable three-dimensional “model” of the building, object, or space. Trained professionals can enter, measure with some reasonable accuracy, and provide relatively quick opinions on the extent of damage and physical quantities of the structure or space they are analyzing. This is done with the same cameras used when taking photographs.
Thornton Tomasetti has also employed the technology to mount the camera to a drone so the photogrammetric model can be obtained from the air, or above locations too dangerous to enter on foot. The speed at which users can generate these models and the ability to document conditions in fuller and more developed imagery is the obvious advantage over static “still” photos. However, there are some limitations; photogrammetry is not as accurate as other “reality capture” methods, such as laser-scanning. Photogrammetry requires post-field processing on a hefty computer to digest and stitch the images together. But, once done, the imagery can be manipulated (and measured) fairly easily. We call this the “poor man’s laser scan.”
Photogrammetry holds the promise of speeding up documentation without the need for costly and time-sensitive laser scanning while being able to be deployed amongst most field technicians (if they are properly trained in taking useful photographs). Even more recently, advances in photogrammetry enable technicians to walk through the site with a HD video camera, input the video into software which can extract still images at regular time intervals, then use those images for photogrammetry input. Naturally, each claim is different and the requirements for accuracy and measurable reliability will differ, but we see this technological advancement improving our ability to evaluate possible damage and provide a fuller capture of conditions at a specific moment in time.
Mark Andrews, Thornton Tomasetti Senior Associate, uses photogrammetry as required for forensic investigations. He shared an interesting segment from NPR where photogrammetry is being used to reconstruct a 3-D models of ancient artifacts and historic sites ruined during violence in the Middle East. “The United Nations reports some 200 sites have been damaged or destroyed by the self-proclaimed Islamic State alone.” Project Mosul, consists of a team of volunteers who digitally reconstruct ancient artifacts from the Mosul Museum in Iraq by using photographs and even video by tourists. These “cyber archeologists” rely on crowdsourcing to upload and sort photos to the project website. To view the digital 3-D renderings, and for the full article, click here.