Apologies if this has already been answered... I couldn't find the answer, though I am sure it exists somewhere.

Take a Kalman Filter for example. If I have a non-linear process model, which I will use either an EKF or a UKF topredict, but I have measurements that can be assumed linear wrt the current state, how do I proceed? Can I use a direct KF for the correction phase?

Thanks

-Phil