Are you using the filter for the barometric sensor? Most digital IMUs now days have self contained filtering mechanisms utilizing multi sensor input (accelerometer, gyroscope, magnetometer)

If this is for the barometric sensor, are you going to be initializing a mach lockout?

I will most likely not be using a filter for the barometric sensor due to the BMP180 is pretty accurate by itself, if I need to institute a filter for it though I would use a simple complementary filter. What I am using this Kalman Filter for is the 9DOF that I get from the Gyroscope, Accelerometer, and Magnetometer due to that they are inherently very noisy and I would like a way of sensor fusion to more accurately control the rocket.

Gotcha. I'm using a complimentary filter for my bmp180, and using a sample offset to calculate simple derivative data for velocity (in conjunction with accelerometer data to insure mach lockout).

I've been pretty happy with my IMU data and it's onboard filtering, but can understand wanting to smooth it out a bit for canard interfacing.
I'm using a BNO055, with onboard sensor fusion cortex-M0, and I'm very happy with the data out of it (fusion data can be accessed at 100Hz)