Research Of Human Motion Capture System Based On Micro-mechanical Sensors

Motion capture is the technology which can measure the movement of moving objectsaccurately in the three-dimensional space, and it is the best way to achieve the rapid creationof three-dimensional animation and the collection of human motion data. Motion capturesystem’s application is more and more widely in the animation. The application of motioncapture technology has solved the bottleneck problem in the process of animation production,and it also has improved the efficiency and quality of animation. This thesis focuses on therelevant technology and the design and implementation of human motion capture systembased on micro-mechanical sensors. To the motion capture system, the collection andtransmission of the human movement data is the most important part, if the system can’t getthe original motion data, and if it will not be able to guarantee the stability and real-time oftransmission., the follow-up study of motion capture cannot be started. This article studies theprinciple of micro mechanical sensors based on model selection and the performance analysisof the sensor. By establishing the model of human body based on dynamics, the systemdesigns sensor nodes which the wireless sensor network contains. The thesis deals with thefunctional analysis and the detailed circuit design of cluster head nodes and sink nodes. Themaster controller to STM32F103CBT6be the core and JN5139wireless transmission modulebased on Zigbee technology, provide the real-time motion data for the system in which updatefrequency is60Hz, and they can guarantee the real-time and stability of the motion datatransmission. At the same time they also have battery detection function. In the part of humanmodel reconstruction software, according to the limitations and constraints of the human bodyjoints, this thesis studies the calibration method of the human body movement posture and therealization process of graphics software. Graphics software can be able to view real-timehuman motion capture animation in all domains. After the completion of every module’sdesign and research, the thesis not only achieves the design of the entire human motioncapture system, but also demonstrates the application of the whole system in the threedimensional animation software Poser7and MotionBuilder. System Demonstration resultsshow that: the function of the system is complete and data transmission is stable.The human motion capture system is mainly involved in sensor technology, thedesigning of sensor node, ZigBee network transmission technology, as well as the design ofhuman motion modeling reconstruction software. It solves the limitation of the way of wiredcommunication. The system gets human gesture data by17micro-mechanical sensors boundat the key joints, after the master controller system capturing, manipulating and sending to thereceiver module, and uploads the data to attitude display software through the USB interfaceof a wireless receiver module. The data captured by sensors needs to be mapped to the virtual characters models. The software inside solvers motion posture of human body through theinverse kinematics algorithm, makes3D body model move, achieves the reappearance ofhuman motion gestures, and conserves the data collected by the system to be applied. Theresearch of the single motion capture system provides some theoretical basis for thedevelopment of multiplayer motion capture system.