Height Estimation for Floor Identification in Elevator and Escalator Scenarios Based on Smartphones Built-in IMU

Accurate floor identification within multi-story buildings is crucial for guiding pedestrians to their destinations in complex door environments like shopping centers, airports and office buildings. Current floor identification methods primarily rely on barometers or wireless signals, which are susceptible to environmental influences or require extensive maintenance. This paper proposes an innovative method for floor identification by estimating pedestrian height using the built-in inertial measurement unit (IMU) of smartphones. By analyzing pedestrian motion in elevator and escalator scenarios, distinct pedestrian motion constraints are derived. An inertial navigation system (INS) framework is employed, which utilizes an extended Kalman Filter (EKF) to integrate these constraints and estimate height changes as pedestrians transition between floors. The method includes detailed motion modeling, taking into account the distinct patterns in elevators and escalators. The proposed algorithm is validated through multiple experiments in a typical shopping mall, demonstrating significant accuracy enhancements with relative height errors of 2.98% (up) and 1.45% (down) in elevator scenarios, and 3.50% (up) and 7.97% (down) in escalator scenarios. This indicates a substantial improvement over existing techniques. Notably, the method does not rely on barometric sensors or prior signal fingerprint databases, highlighting its potential universality across common mobile devices.