Telemedicine applications, based on twodimensional(2D) video conferencing technology, have been around for thepast 15 to 20 yr. They have been demonstrated to be acceptablefor facetoface consultations and useful for visual examinationof wounds and abrasions. However, certain telerehabilitationassessments need the use of spatial information in order to accuratelyassess the patient’s condition and sending threedimensionalvideo data over lowbandwidth networks is extremelychallenging. This article proposes an innovative way of extractingthe key spatial information from the patient’s movementduring telerehabilitation assessment based on 2D video and thenpresenting the extracted data by using graph plots alongside thevideo to help physicians in assessments with minimum burdenon existing video data transfer. Some common rehabilitationscenarios are chosen for illustrations, and experiments are conductedbased on skeletal tracking and color detection algorithmsusing the Microsoft Kinect sensor. Extracted data are analyzedin detail and their usability discussed.
展开▼