Fundamentals of Wearable Computers and Augmented Reality, 2nd Edition

220 Fundamentals of Wearable Computers and Augmented Reality

The Kalman ltering process for orientation estimation can be broken down into

two primary steps: the prediction step and the updating step. In the predicting phase,

the lter uses the gyroscope measurement to predict the dynamics of the device rota-

tion. The gyroscope measurements are thus integrated directly into the state transi-

tion equation and used to provide a predicted state estimate. In particular, the state

equation is dened using a seven-element state vector consisting of the quaternion

qt()

(i.e., four-element orientation vector) and the gyro-bias b(t):

Xt

qt

bt

()

()

()

=

