I'm implementing a kalman filter that fuses data from 2 different sensors. Both sensors provide the same data (a 3d position measurement), but the variances for each sensor are different (i.e. one is more accurate than the other).
From the resources I've read (for example, here and here), I understand that the measurement noise covariance matrix is a matrix whose elements are the degree of correlation between the ith state variable and the jth state variable. Generally, this ends up being a diagonal matrix whose elements are the variance. I understand how this would work if both of my sensors had the same characterization. How can I account for my sensor differences in the measurement noise covariance matrix? Or, am I misunderstanding how the matrix works?
I'm having trouble finding examples and resources for my use case (i.e. sensor fusion of like data), so any help or direction is very much appreciated. Thank you!