Research And Implement Of Interactive Body Training System

In recent years,rehabilitation training robots have gradually replaced the traditional rehabilitation training method,which can help patients to carry out accurate,automatic and intelligent rehabilitation training,and then they have become a safe and effective rehabilitation method.This paper analyzes the requirements of a kind of human-computer interaction software applied to mobile rehabilitation training robot and operating desktop.The main research work is as follows,(1)Research on the scheme of interactive body training system.Based on the requirement analysis of interactive body training system,the overall scheme of the software system,the main functional modules in the system and the interaction between modules are designed.The concept design and data table design of the database of the software system are carried out to complete the effective management of patients’personal information and rehabilitation information.(2)The design of interactive body training system.The content and interface of the main function modules of the system are designed.The virtual interactive model of mobile rehabilitation training robot is created in the interactive module of lower limb mechanical training.The information transmission between the virtual interactive model and rehabilitation training robot is completed through the data communication module,and then the real motion state is visualized.(3)The design of lower limb virtual interactive training module.According to the functional characteristics and training requirements of the mobile rehabilitation training robot,different scene nodes are arranged in the virtual scene,and the corresponding interactive tasks are designed to guide patients to complete a variety of postures of lower limb training,which increases the interest of patients’ rehabilitation training.(4)The design of hand grasping interactive training module based on monocular vision.Under the monocular camera,the hand contour area,virtual feature parameters of length and virtual feature parameters of area are defined to evaluate the grasping state of the hand by analyzing the features of the hand in flat,vertical and inclined posture.The real-time interaction between the hand and the virtual model is realized based on the change of parameters.At the same time,a method to recognize the bending angle of the metacarpophalangeal joint is proposed,which realizes the real-time interaction with the virtual model in the process of joint bending.