I'm trying to track an accelerating vehicle using a camera, an IMU, and a GPS. I use for the state space equation a constant acceleration model: The states are the position, the velocity, and the acceleration of the vehicle in the x- and y-direction. No input is used. All the measurement equations are linear. I'm wondering if I should use a discrete normal Kalman filter, or a discrete extended Kalman filter. (i.e. is my process non linear?)
If you can write the dynamics with a matrix, which you have, then a normal kalman filter will be best.
However, your measurements will probably be nonlinear. You will find that you won't be able to write your measurements equations using matrices. You will almost certainly need an extended kalman filter because your measurements will be nonlinear.