A review of computer vision-based approaches for physical rehabilitation and assessment

Abstract The computer vision community has extensively researched the area of human motion analysis, which primarily focuses on pose estimation, activity recognition, or gesture recognition and so on. However for many applications, like monitoring functional rehabilitation patients with musculo skeletal physical impairments, requirement is to comparatively evaluate motion. In this survey, we capture important literature vision-based that comparative evaluation during past two decades discuss state current research in area. Unlike other reviews area, are written from a clinical objective, article presents application perspective. We propose our own taxonomy assessment further divided into sub-categories novelties each research. review discusses challenges domain due wide ranging abnormalities difficulty automatically assessing those abnormalities. Finally, suggestions future direction offered.