Riassunto analitico
In human motion tracking, estimating the orientation of human body plays an important role. Since commercial sensors generate noisy measurements, computationally efficient algorithms for mitigating the impact of noise are required. In this thesis, a quaternion-based extended Kalman filter for determining the orientation of a rigid body using tri-axis sensing based on accelerometer, gyroscope and magnetometer is developed. Our aim is implementing a Kalman filter to mitigate noise affecting the measurements provided by an inertial measurement unit (IMU) and to provide an accurate orientation of a rigid body. More specifically, the method developed in this work is based on an extended Kalman filter (EKF) and on the use of quaternions. This method is implemented in the MATLAB environment.
|
Abstract
In human motion tracking, estimating the orientation of human body plays an important role. Since commercial sensors generate noisy measurements, computationally efficient algorithms for mitigating the impact of noise are required. In this thesis, a quaternion-based extended Kalman filter for determining the orientation of a rigid body using tri-axis sensing based on accelerometer, gyroscope and magnetometer is developed. Our aim is implementing a Kalman filter to mitigate noise affecting the measurements provided by an inertial measurement unit (IMU) and to provide an accurate orientation of a rigid body. More specifically, the method developed in this work is based on an extended Kalman filter (EKF) and on the use of quaternions. This method is implemented in the MATLAB environment.
|
