Open AccessForensic Engineering Usage of Surveillance Video in Accident Reconstruction
Author(s) -
Richard M. Ziernicki,
William H. Pierce,
Angelos G. Leiloglou
Publication year - 2014
Publication title -
journal of the national academy of forensic engineers
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.102
H-Index - 1
eISSN - 2379-3252
pISSN - 2379-3244
DOI - 10.51501/jotnafe.v31i2.2
Subject(s) - computer science , computer vision , video quality , distortion (music) , artificial intelligence , computer graphics (images) , frame (networking) , engineering , telecommunications , amplifier , metric (unit) , operations management , bandwidth (computing)
With the increased use of surveillance cameras, more and more video footage depicting accidents is available these days for accident reconstruction. The authors present an accident reconstruction case study involving an impact between a tractor-tanker and a pedestrian using surveillance video footagefrom a nearby business. Overall, the video footage is of poor quality, which is typical of surveillance video. This is usually evidenced by low frame rate, low resolution, and significant lens distortion not to mention the fact that the video is not centered on the actual accident. This paper addresses a solution to minimize the error often associated with such surveillance video. First, the distortion in the video footage is corrected using software that warps the image with a reverse distortion. Once the distortion in the video footage is corrected, then accurate photo/videogrammetry is performed to attain desired measurements. These measurements are then processed to perform a more accurate and detailed time/space analysis. Finally, graphics and photo-realistic animation are used to present the accident in time-space domain.