Indoor positioning based on lower limb motion

Indoor positioning provides great convenience for people's daily life. So far, numerous techniques have been proposed to track human movement, such as TOA, AOA and RSSI. Usually, these methods measure position depending on installed cells. In this paper, we propose a lower limb motion-based indoor positioning approach. Four IMU sensors are attached on both legs, separately, in order to get lower limbs orientation during a walking action. Step alternations are obtained by detecting the supporting foot, with the assistance from virtual colliders in re-constructed virtual platform. Furthermore, while the supporting foot is moving back and forward, a compensatory displacement is calculated to represent the actual displacement of human body. Finally a transformation between the global and Unity3D coordinate systems works out successive indoor positions. Stability and reliability are verified by in-situ walking tests. Indoor walking experiments show 9.9 cm position drift and 38.7 cm
²
variance over a 700 cm long route. Without overdependence on beforehand installed cells of TOA's or RSSI's, our method indeed provides a novel way of acquiring human positions and it extends the applications of motion capture, at the same time.