3D indoor crime scene reconstruction from micro UAV photogrammetry technique

• Main Authors: Sazaly, Alysa Nur, Mohd. Ariff, Mohd. Farid, Razali, Ahmad Firdaus
• Format: Article
• Language: English
• Published: Dr D. Pylarinos 2023
• Subjects: G Geography (General)
• Online Access: http://eprints.utm.my/106851/1/MohdFaridMohdAriff2023_3DIndoorCrimeSceneReconstruction.pdf; http://eprints.utm.my/106851/; http://dx.doi.org/10.48084/etasr.6260

The application of micro Unmanned Aerial Vehicles (UAVs) in photogrammetry, particularly within the realm of forensic investigation represents a relatively novel approach and has gained increased attention. By measuring the distances and positions of the scene’s components, it is feasible to document and visualize the scene using the photographs that were taken for the purpose of assisting investigators. Capturing accurate crime scene data within a short time frame is always a challenge. Conventionally, photographs were used to document the scene, but the technical qualities of the photographs depended on the skill of the present forensic personnel. The use of 3-Dimensional (3D) photogrammetry enables the production of highly realistic and detailed 3D documentation of a given scene. As this technique involves capturing a series of photographs, it can be a time-consuming process. Therefore, this study aims to explore an alternative approach that enables the rapid acquisition of the scene while preserving the intricate details, thus ensuring efficiency without compromising the accuracy of the resulting documentation. The study employs a methodological approach wherein data are collected from a simulated crime scene situated within a confined and hard-to-reach area. The data collection is facilitated through the utilization of micro UAVs. The acquired data are then processed utilizing photogrammetry software, leading to the generation of a 3D model point cloud. The collected data will be subjected to a comparative analysis with data generated using a Terrestrial Laser Scanner (TLS) as a reference, alongside Vernier Calliper (VC) measurements. The findings indicate that the Root Mean Square Error (RMSE) of the integrated point clouds from TLS and micro UAVs compared to the conventional method is approximately ±0.217 cm. It can be deduced that the integration of data derived from micro UAVs and TLS in forensic photogrammetry within a confined crime scene is viable and yields a high-precision 3D model point cloud.