Camera System for Efficient non-contact Measurement in Distance Medicine
Patrik Kutílek, Jan Hejda, Lenka Lhotská, Jindřich Adolf, Jaromír Doležal, Michaela Hourová, Pavel Kral, Yoram Segal, Raz Birman, Ofer Hadar · 2020
The posture of body segments can be negatively influenced by many diseases of the nervous, visual and musculoskeletal systems. This article outlines a newly designed system and related procedures to record and evaluate anatomical body angles. The system is equipped with two mutually calibrated cameras allowing the evaluation of body movement in two anatomical planes. The hardware part of the camera system and calibration method was designed for practical use in clinical practice. Moreover, the proposed camera system allows for the recording and evaluation of motion data in a home environment or at a safe distance. It enables non-invasive and non-contact measuring of body segments and, therefore, can be used in distance rehabilitation and distance diagnosing. The study also demonstrates the hardware's performance accelerator based on a human body tracking algorithm. The device utilizes the third party algorithms, such as OpenPose, for the extraction of major body points from selected video frames. Any missing data points are then interpolated through the proposed tracking algorithm. This procedure results in an acceleration of the overall hardware performance.