Enhancing Physical Rehabilitation Monitoring through Multi-View Fusion of Human Body Skeleton

The Kinect has gained popularity in physical therapy and rehabilitation systems as a skeleton body tracking device. However, limitations such as occlusion and noise in skeleton tracking have reduced its reliability. To address these issues, some studies have proposed multi-kinect systems to expand the field of vision and enhance data precision. In this paper, we introduce the design of a 360° multi-view fused human body skeleton system using Kinect devices. Our system harnesses the strengths of the Kinect's Software Development Kit (SDK) and the spatial arrangement of the Kinect devices to estimate the 3D position coordinates of human skeleton joints. Importantly, the system allows for easy customization to accommodate modifications. Our system enables the extraction of indicators related to human walking patterns within the context of physical rehabilitation. These indicators serve to motivate patients, provide valuable feedback, and track their rehabilitation progress. The system demonstrated its utility in initial laboratory tests for physical therapy and rehabilitation monitoring. Our research presents a novel approach to address the limitations of the Kinect in physical therapy and rehabilitation. The preliminary findings from laboratory tests indicate the system's potential to enhance physical therapy and rehabilitation monitoring.