Examination of fire scene reconstructions using virtual reality to enhance forensic decision-making. A case study in Scotland.

Vincenzo Rinaldi,Karen Ann Robertson,Graham George Strong,Niamh Nic Daeid

DOI: https://doi.org/10.1007/s10055-024-00961-w

IF: 4.697

2024-03-02

Virtual Reality

Abstract:When attending a crime scene, first responders are responsible for identifying areas of potential interest for subsequent forensic examination. This information is shared with the police, forensic practitioners, and legal authorities during an initial meeting of all interested parties, which in Scotland is known as a forensic strategy meeting. Swift documentation is fundamental to allow practitioners to learn about the scene(s) and to plan investigative strategies, traditionally relying on word-of-mouth briefings using digital photographs, videos, diagrams, and verbal reports. We suggest that these early and critical briefings can be augmented positively by implementing an end-to-end methodology for indoor 3D reconstruction and successive visualisation through immersive Virtual Reality (VR). The main objective of this paper is to provide an integrative documentation tool to enhance the decision-making processes in the early stages of the investigation. Taking a fire scene as an example, we illustrate a framework for rapid spatial data acquisition of the scene that leverages structure-from-motion photogrammetry. We developed a VR framework that enables the exploration of virtual environments on a standalone, low-cost immersive head-mounted display. The system was tested in a two-phased inter-agency fire investigation exercise, where practitioners were asked to produce hypotheses suitable for forensic strategy meetings by (1) examining traditional documentation and then (2) using a VR walkthrough of the same premises. The integration of VR increased the practitioners' scene comprehension, improved hypotheses formulation with fewer caveats, and enabled participants to sketch the scene, in contrast to the orientation challenges encountered using conventional documentation.

computer science, interdisciplinary applications, software engineering,imaging science & photographic technology