I am looking to improve gps tracking for an app that i inherited. We track latitude, longitude and altitude. After tracking we give stats on speed, distance and vertical descent. The code currently uses a low pass filter to keep the altitude in check, but does nothing with lat lon. Which causes issues around speed, distance.
I was looking at implementing a Kalman filter to help reduce noise. My question is around the different versions of the Kalman filter. With what we need to track can I get by with a Kalman filter or do I need to investigate an extended / unscented Kalman filter?
Thanks
You will need an extended kalman filter because the physical model for an object moving is non-linear.
It can suffer accelerations, direction changes etc, so the kalman filter will not be able to track it properly. Tuning an extended kalman filter is more tricky, but you can find quite a lot of papers and information about that.