Skip to content

3D Gaussian Splatting Offers New Possibilities for Forensic Autopsy Documentation

New research examines how 3D Gaussian Splatting can improve forensic autopsy documentation through optimized photographic acquisition

By 2 min read

Three-dimensional imaging is gaining attention in forensic science as researchers explore ways to improve the documentation and digital examination of forensic findings. A new study published in the Journal of Forensic Sciences investigates how 3D Gaussian Splatting (3DGS) can be used to create detailed visual reconstructions from photographs collected during forensic autopsy documentation.

The research, titled 3D Gaussian Splatting for Forensic Autopsy Documentation: Influence of Image Acquisition Strategy on Reconstruction Quality, was conducted by Constantin Lux, Florian Kreitmeier, Wolf Schweitzer and colleagues from institutions in Switzerland and Germany.

3D Gaussian Splatting is an image-based reconstruction technique that represents a three-dimensional scene using numerous Gaussian elements. It allows users to generate photorealistic views and interactively examine reconstructed objects or environments from different angles.

The researchers investigated ten forensic autopsy cases using a standardized photographic documentation workflow. They compared three image-acquisition strategies: region-based multi-angle acquisition (RMA), fixed-angle circumferential acquisition (FCA) and continuous angular acquisition (CAA).

To evaluate reconstruction quality, the researchers compared rendered images with reference photographs using three established image-quality measures: Peak Signal-to-Noise Ratio (PSNR), Structural Similarity Index Measure (SSIM) and Learned Perceptual Image Patch Similarity (LPIPS).

The results showed significant differences between the acquisition strategies. Continuous angular acquisition consistently achieved higher PSNR and SSIM values and lower LPIPS scores than the other two approaches, indicating better similarity between reconstructed images and their photographic references.

The researchers also identified reconstruction artefacts in anatomically complex and weakly textured regions. Vignetting correction, which compensates for variations in image brightness, improved texture stability and helped reduce these artefacts.

The findings highlight an important factor in forensic imaging: reconstruction quality depends not only on the software used but also on how photographs are captured and processed.

However, the study has an important limitation. Its evaluation relied on image-based quality metrics, which assess visual similarity rather than geometric accuracy or the reliability of physical measurements. Therefore, 3DGS should currently be understood as a complementary visual documentation method rather than a replacement for established measurement-based forensic techniques.

The research provides practical evidence that carefully planned image acquisition can improve 3D forensic documentation and supports further investigation into the use of advanced visualization technologies in forensic practice.