Indoor Navigation System Based on Foot-Mounted IMU and Map Information Fusion

An efficient correction method for position and heading errors in foot mounted inertial measurement unit (IMU) indoor pedestrian navigation and positioning systems is presented in this paper. We propose a cascaded structured Kalman filter and non-recursive Bayesian filter for indoor localization by fusing the micro electro mechanical system (MEMS) inertial sensors and indoor map information. The lower Kalman filter adopts the zero velocity update algorithm to initially correct the solution error of inertial navigation; the upper non-recursive Bayesian filter uses indoor map information to further calibrate pedestrian position and heading by map matching. The effectiveness and accuracy of this algorithm are verified by example in a real indoor scene.