Pedestrian detection with high resolution inertial measurement unit

Inertial sensors are used widely for detecting different contexts. However, noise components at high frequencies can disturb the recognition. In this paper, the measurements are made with multi IMU (MIMU) which combines the results of 32 inertial sensors. Averaging of individual sensor outputs reduce the noise level significantly and enables higher resolution. As an example case, we present application for passenger detection in two environments; hallway corridor and public city bus. The results show that accuracy can be increased when MIMU is used compared to single IMU.