Pixel Binning Effects of Smartphone Camera on Three-Dimensional (3D) Model Reconstructed Crime Scene

Pixel binning, a feature of high-megapixel smartphone cameras, exhibits performance comparable to traditional cameras. The field of photogrammetry has explored and adopted most kinds of technology, hence, pixel binning too should be adopted into forensic photogrammetry. This study evaluates the application of pixel binning technology in forensic photogrammetry, specifically in 3D reconstruction at crime scenes. A simulated crime scene conducted at the UTM-PDRM lab was captured using smartphone cameras of 12MP and 50MP, and a 20MP DSLR camera. First, the cameras were calibrated to ensure their stability. Following the image capture, the data were processed to generate 3D point cloud models of the simulated crime scene. The geometric parameters resulting from the camera calibration were discussed. The 3D point cloud model by DSLR camera exhibited better visual quality than the smartphone cameras. This finding was supported by an analysis of overlapping images by each camera and a side-by-side comparison of the models. Measurements from the smartphones 1, 2 and the DSLR camera were compared to conventional Vernier calipers used in crime scene documentation. The resulting Root Mean Square Error (RMSE) differences were approximately ±5.62mm, ±5.59mm, and ±5.40mm, respectively. In conclusion, the pixel binning of smartphone cameras was able to produce reliable accuracy but requires stability in technology for 3D reconstruction.