This paper presents a fast six-position calibration for strapdown inertial measurement unit (IMU) bias. The
interference noise is filtered out with hidden Markov model/ Kalman filter (HMM/ KF), and the alignment is achieved based
on traditional analytic coarse alignment. The mathematical expressions of bias is derived with the relationship between
the accuracy of coarse alignment and inertial instrument errors. Compared with the traditional coarse alignment, the initial
alignment with filter not only can improve azimuth accuracy, but also can compensate the residue bias in the preceding
calibration period, and navigation performance is improved.