IoT Framework for 3D Body Posture Visualization

Visual feedback is a powerful tool that can assist in both the training and recovery processes. During training, athletes may correct poor posture or identify potentially hazardous movements. Likewise, physicians may be able to identify postures that would lead to further injury. Additionally, such a system could also be beneficial during fall detection to provide greater insight into the patient's position and status. Current models for providing feedback to the user rely on full-body sensor sets or video representations, which may cause discomfort or may not fully capture the user's motion. We propose a new system architecture that we will define as a 3D-BPV (Body Posture Visualization) system. This paper seeks to design a less intrusive sensor system, based on Internet of Things (IoT) technology, which visualizes patient movement in a 3D model. A Kalman Filter will also be used to eliminate sensor drift during operation. The system should minimize the size and number of sensors attached to a patient while providing sufficient data for generating such a model. To demonstrate such a design, a system using accelerometers has been constructed with the 3D model generation accomplished using a Biovision Hierarchy Animation (BHV) file.