Real-time point cloud data transmission in Kinect point cloud data based telerehabilitation system

Real-time monitoring is a key aspect of Telemedicine. Using Kinect sensor for remote video based rehabilitation diagnosis has been a low cost, convenient, and effective technology. However, point cloud data from Kinect sensors are usually quite large (90MB/s). Therefore, how to ensure the quality of service (QoS) of data transmission under limited bandwidth is a challenging issue. In this thesis, we propose a self-adaptive and hybrid mode with both patient self-management and doctor remote diagnosis based telemedicine framework. The core of this framework is a Kinect differential octree based point cloud data dynamic algorithm, point cloud detail compression, frame skipping, dynamic compression and etc. We performed numerical analysis and simulation. Results show the system has good effective and feasibility under limited bandwidth. In addition, a Speed-based Telerehabilitation System was proposed as test system for dynamic algorithm.