Low-Cost 3D Vision Based System for Post-Stroke Arm Rehabilitation using RGB-D Camera

• Autor(a) principal: Al-Kaff, Abdulla
• Data de Publicação: 2016
• Tipo de documento: Artigo
• Idioma: eng
• Título da fonte: INFOCOMP: Jornal de Ciência da Computação
• Texto Completo: https://infocomp.dcc.ufla.br/index.php/infocomp/article/view/518
• topic: Markerless Motion Capture; Kinect Sensor; Virtual Reality; Arm Rehabilitation

In this paper, a low-cost markerless system based on RGB-D sensor was developed for rehabilitation purposes.The purpose of this system is to provide the physical therapist with the adequate data to characterize the arm motion in post-stroke rehabilitation, athletic injuries recovery or other possible human arm motion recording and analysis issues. The proposed system uses the Microsoft Kinect sensor to determine the motion of the patient arm.The 3D position of the joints (shoulder, elbow and wrist) and the centers of mass of arm segments (upper and forearm) are estimated during the motion. The system then calculates the velocities and energies expenditure by each segment of the arm.More data are estimated by the system such as the angles of each joint (shoulder and elbow) and the angular velocities of the arm segments, where these data are useful to help the therapist with a full overview in the analysis about the patient motion.With the help of these data, the therapist has the analysis of the arm, and comparing with the previous analysis of the patient arm motion, a relative performance can be estimated, allowing a quantitative evaluation of the recovery of the patient over series of exercises.Furthermore, as a part of the work, 3D Virtual Reality System $(VRS)$ was developed. This 3D VRS can be a beneficial environment for learning a motor task and it is important in terms of the motion motivation, since it presented to the patient as a rewarded computer game.