A Depth-Camera Based System for the Real-Time Scanning of Upper Limb Anatomy

The 3D reconstruction of upper limb anatomy plays a significant role in many biomedical fields as ergonomics, motion rehabilitation, prosthesis design. Conventional manual measurements have been progressively replaced by 3D optical scanning in collecting and storing 3D anatomical data, thus increasing reliability and data accuracy, shortening, at the same time, the overall acquisition process. However, the real-time scanning of human body parts still represents a complex task since it is challenging to keep the arm in a stable position and avoid artifacts in the collected data. Also, optical undercut geometries often impair the 3D reconstruction’s completeness. In this paper, a compact and low-cost 3D scanning system has been developed by integrating three D415 Intel RealSense cameras. The three depth cameras have been assembled in a circular rig to define a lightweight handheld scanner capable of carrying out 3D data acquisition in different scenarios. The optical system has been validated through anthropometric measurements on different subjects.