Geometric Accuracy Investigations of Mobile Phone Devices in the Laboratory Using High-Precision Reference Bodies
Keywords: 3D, accuracy, comparison, dense point cloud, meshed model, reference data, smartphone
Abstract. The sensors in modern mobile phones (running either the Android or iOS operating system) have become increasingly sophisticated, to the extent that they can be used as measuring systems for a wide range of applications. On the one hand, GNSS (Global Navigation Satellite Systems) and IMU (Inertial Measurement Unit) provide precise positioning of smartphone sensors. On the other hand, the inbuilt cameras offer an increasingly high geometric image resolution. In order to investigate the potential of mobile phones for creating 3D models of small objects for the documentation of museum artefacts, the Laboratory for Photogrammetry & Laser Scanning of the HafenCity University Hamburg tested various smartphones for geometric accuracy under laboratory conditions. Four Galaxy S-series smartphones of Samsung (S21+, S22, S23, S24 Ultra) and two Apple iPhones (13 Pro and 15 Pro Max) were used for the tests. The image data sets of three distinct test objects, captured with disparate mobile devices, were processed in Agisoft Metashape into 3D models by triangle meshing and subsequently compared with highly accurate reference data from an ATOS 5 structure-light projection system. Some selected examples of image data sets recorded with the iPhone 15 Pro Max were also processed in the Polycam app. The results of the geometric accuracy analyses demonstrated that the image data captured by smartphone cameras could be processed into highly accurate three-dimensional models of the objects. The deviations from the reference data were only marginally inferior to those observed in the models generated from image data obtained from a SLR camera.