Public Safety Implementation of Unmanned Aerial Systems for Photogrammetric Mapping of Crash Scenes

Abstract

Terrestrial photogrammetry using acquired images by a hand-held camera has been used for several years to map crash scene geometry. More recently, photogrammetric reconstruction from acquired images by an unmanned aerial system (UAS) has been proposed for crash scene mapping. Over the past year, the Tippecanoe County Sherriff’s Office has participated in three workshops with Purdue University, applied these skills in two training mass casualty exercises, and independently mapped five crash scenes in June and July 2018. This paper briefly reviews the training sessions, mass casualty exercises, and five crash scenes mapped by Tippecanoe County Sherriff’s deputies. The paper presents a comparison of both traditional ground-based and UAS-based photogrammetric mapping for two crashes in July 2018. The UAS procedures described in this paper are quite similar to current ground-based photogrammetric mapping. The UAS-based photogrammetric mapping derived measurements from eight identified crash scene markers and key features were found to be within 0.29 ft of field tape measurements, or with 0.4% or less relative error and a root mean squared error of 0.12 ft. We believe this paper will become important documentation in the literature that will provide public safety agencies with performance data to support their deliberation in investing in this new technology.