Format

Send to

Choose Destination
Sensors (Basel). 2018 Mar 1;18(3). pii: E741. doi: 10.3390/s18030741.

A Foot-Mounted Inertial Measurement Unit (IMU) Positioning Algorithm Based on Magnetic Constraint.

Author information

1
School of Computer Science and Technology, China University of Mining and Technology, Xuzhou 221116, China. wystephen@cumt.edu.cn.
2
School of Computer Science and Technology, China University of Mining and Technology, Xuzhou 221116, China. linuxcumt@126.com.
3
School of Environmental Science and Spatial Informatics, China University of Mining and Technology, Xuzhou 221116, China. zoujiaheng@cumt.edu.cn.

Abstract

With the development of related applications, indoor positioning techniques have been more and more widely developed. Based on Wi-Fi, Bluetooth low energy (BLE) and geomagnetism, indoor positioning techniques often rely on the physical location of fingerprint information. The focus and difficulty of establishing the fingerprint database are in obtaining a relatively accurate physical location with as little given information as possible. This paper presents a foot-mounted inertial measurement unit (IMU) positioning algorithm under the loop closure constraint based on magnetic information. It can provide relatively reliable position information without maps and geomagnetic information and provides a relatively accurate coordinate for the collection of a fingerprint database. In the experiment, the features extracted by the multi-level Fourier transform method proposed in this paper are validated and the validity of loop closure matching is tested with a RANSAC-based method. Moreover, the loop closure detection results show that the cumulative error of the trajectory processed by the graph optimization algorithm is significantly suppressed, presenting a good accuracy. The average error of the trajectory under loop closure constraint is controlled below 2.15 m.

KEYWORDS:

FFT; foot-mounted IMU; graph optimization; magnetic field

Supplemental Content

Full text links

Icon for Multidisciplinary Digital Publishing Institute (MDPI) Icon for PubMed Central
Loading ...
Support Center