Research And Implementation Of The Wireless Sensor Network-based Limb Motion Capture Technology

The popularity of Limb Motion Capture technology is increased dramatically because of its wide application. Traditional solutions have many limitations such as high equipment cost and complexity, and stringent environmental quality requirement. Based on Wireless Sensor Network, the advantage of using MEMS(Micro-Electro-Mechanical Systems)sensor networks to collect data and inertial navigation technology to process it lies in smaller size of equipment, lower cost, higher independency and so on. Utilizing wireless network to transfer data cuts the device size and simplifies connections, which results in reduced physical burden. Therefore, there would be broader prospect of development for Limb Motion Capture technology based on Wireless Sensor Network.Wireless Sensor Network based on ZigBee protocol is analyzed in this thesis. The wireless senor network is developed based on Z-Stack protocol stack platform. Three-axis acceleration sensor ADXL345 and gyroscope L3G4200 D on sensor nodes are utilized to collect limb movement data, and Wireless Sensor Network is used to realize data transmission and aggregation. Data is then processed by inertial navigation technology. Angular velocity data is used to correct the direction of acceleration. Moreover, the acceleration data referenced by the coordinate system of the sensor is mapped to the geographic coordinate system. By integrating the acceleration data to get the velocity and displacement of each node, the trajectory of each joint of limb is obtained, and eventually,the limb movement state.The designed system utilizes hardware compensation to reduce the sensor error. Kalman filter algorithm is adopted to reduce the impact of accidental error and low dead area algorithms to process cumulative error to increase the data accuracy. By comparing errors under different experiment conditions, it is conclued that the error-handling method put forward in this thesis works effectively, which can be used to draw a simple picture of limb movement accurately.