This paper presents the development of an Extended Kalman Filter (EKF) that optimally processes photographic data collected during short-duration impact acceleration tests. The system is modeled by a non-linear state-space representation using quaternions for rotational kinematics. Three cameras are used to photograph up to 14 fiducials mounted on a plate attached to the subject's mouth. The filter yields the history of the rotational and translational kinematics of the origin of the mouth plate. Results from the EKF and analysis of the estimation error are presented.
Kaminsky, E. J., R. E. Trahan, P.M. Chirlian, and B. King, "Extended Kalman Filter for Photographic Data from Impact Acceleration Tests," lASTED Signal and Image Process. Cont. Proc. (SIP'96), Orlando, FL, Nov. 11-14, 1996, pp. 59-63.