Design And Research Of An Online Rehabilitation Training State Recognition System

Records of patients' articulation data in rehabilitation training are significant.Joint node data method based on the visual capture technique is a new means of data collection.This method is both economical and practical,with a broad space of development.Based on the motion joint data collected by Kinect 2.0 equipment,this paper designs and analyzes an online rehabilitation training status recognition system based on the error in the modified data and realizes the real-time visualization of the data.It can reflect the image information of rehabilitation training as well as the trajectory of lower extremity joint.The system helps to improve the existing rehabilitation data recording model and improve the efficiency and accuracy of rehabilitation data records.This paper summarizes the current situation and existing problems at home and abroad,and expounds the necessity and significance of recording the rehabilitation training data of patients and the key problems to be solved.This paper analyzes the error and cause of the trajectory data of the ankle and knee joint collected in the rehabilitation exercise.A fast,simple and real-time neural network trajectory correction algorithm is designed as the core algorithm of rehabilitation training state recognition system.On this basis,the system module planning and programming are carried out.Then,the overall design of the online rehabilitation training state recognition software system is completed.The online rehabilitation training status recognition system was tested.