Research On Real-time Rendering Technology For Assessment And Training Of Ar-Based Elbow Rehabilitation

Augmented Reality(AR)is a technology that seamlessly fuses real-world information with computer-generated virtual information.AR technology is applied to assessment and training of elbow rehabilitation,it can seamlessly superimpose computer-generated virtual information on the patient's actual rehabilitation training process,and it provides an interesting,stable,and personalized training method.The research on real-time rendering technology for AR-based elbow rehabilitation system of assessment and training is a further study of real-time rendering methods and applications,it provides a theoretical basis for the application of AR technology on the rehabilitation field and gives key technical solutions for assessment and training system of AR-based elbow rehabilitation.Therefore,the work of this thesis has important theoretical significance and application value.Real-time rendering is one of the key technologies of AR,and it is also the basis for virtual-real fusion.Currently,AR-based elbow rehabilitation systems of assessment and training are relatively few and not targeted at home and abroad.There is a problem of low tracking robustness and registration accuracy based on general marker,and it is difficult to balance realism and real-time of virtual model rendering.In view of the above issues,the main work of this thesis based on the in-depth research and analysis of related AR real-time rendering theory and methods is as follows:(1)Analysis of virtual-real fusion.The location of the marker is determined by the analysis of characteristics of the elbow activity.The flexion,extension and pronation,supination of elbow can be abstracted as rotation in a three-dimensional space around a coordinate axis,so that the marker worn on the forearm is mathematically modeled in three-dimensional space,and then the virtual object rendering position is determined.A real-time rendering method suitable for the mobile terminal is determined combined with the existing real-time rendering technology.(2)The improved marker and rendering algorithm is suitable for the elbow rehabilitation of assessment and training.The marker is easy to wear and easy to accept because of its small volume and low weight.In view of the traditional marker tracking drawbacks of low robustness,this thesis selects a color marker combining feature points extraction algorithm to solve the problem that the marker is unstable when it follows the elbow motion.The geometric consistency of the virtual model is guaranteed at the same time.The method of model rendering combined with local illumination model and Gouraud algorithm solves the problem of low real-time rendering of the virtual model on the premise of ensuring a high degree of realism and illumination consistency.The experiment verifies stability and real-time performance of this method.(3)Assessment of effectiveness and availability.On the one hand,a prototype is designed and developed by oneself,and the experiments are carried out in hospital.The healthy,test patients and real patients are selected as evaluation targets.A set of experiments is designed to analyze the effectiveness of the rendering results.On the other hand,the effectiveness of real-time rendering is further verified from the availability of hardware(mainly AR marker)and software.Hardware usability is evaluated from four aspects: convenience,fault tolerance,comfort and acceptability,and software availability is evaluated in terms of efficiency,attractiveness,fault tolerance,easy to learn and acceptability.The innovative points of the thesis are mainly reflected in the following:(1)The improved marker combined with the characteristics of elbow movement is used for abstract modeling and completing stable tracking and registration.Real-time rendering is achieved combined with improved model rendering.(2)The real-time rendering method realized by the packaging application in AR-based elbow rehabilitation of assessment and training prototype system not only provides a virtual-real training situation for patients,but also can be used as an important basis for doctors to evaluate the patients' rehabilitation process.In addition,the work of this thesis is a combination of AR field and rehabilitation medicine field,the results provide a theoretical basis and reference for elbow rehabilitation of assessment and training and balance the stressful medical resources at the same time.This thesis achieves real-time rendering for assessment and training of elbow rehabilitation based on improved marker.The next task is to support evaluation and training of elbow rehabilitation better by tracking registration and real-time rendering based on the direction sensor.