Transforming Physical Crime Scene into Geospatial-based Point Cloud Data

Abstract

Terrestrial Laser Scanning (TLS) and Close-Range Photogrammetry (CRP) are advanced techniques for capturing 3D data in crime scene reconstruction, offering complementary information. Despite taking multiple scans and images from different angles to ensure a comprehensive model, limitations, such as device positioning, shadows, object distance, and laser beam angles prevent the creation of a complete crime scene model. Therefore, combining TLS and CRP data is crucial for achieving a comprehensive reconstruction. This study aims to transform a physical crime scene into a geospatial-based reconstructed model known as point clouds. The technique used was highly rich in realistic features, digitally reconstructed from TLS and CRP. The data sources were then fused via a rigid body transformation, creating a comprehensive crime scene model. The combined point cloud measurements were compared with measurements obtained from a high-precision Vernier caliper to ascertain their accuracy. The resulting Root Mean Square (RMSE) difference between the fused point cloud data and the high-precision caliper measurements was approximately ±4mm. The fusion of TLS and CRP data provides reliable and highly accurate 3D model point clouds, making it suitable for forensic applications.