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

Area of Science:

• Robotics and Navigation
• Geospatial Information Science
• Sensor Fusion

Background:

• Indoor positioning systems (IPS) commonly utilize Wi-Fi, Bluetooth low energy (BLE), and geomagnetism, often requiring fingerprinting.
• Establishing accurate fingerprint databases is challenging due to the need for precise physical location data with minimal prior information.
• Existing methods face difficulties in achieving high accuracy and reliability, especially in environments lacking detailed maps or extensive sensor networks.

Purpose of the Study:

• To develop a novel indoor positioning algorithm for creating accurate fingerprint databases.
• To enhance the reliability and accuracy of indoor positioning using a foot-mounted inertial measurement unit (IMU).
• To validate the effectiveness of loop closure constraints and magnetic information in reducing cumulative trajectory errors.

Main Methods:

• Implementation of a foot-mounted inertial measurement unit (IMU) positioning algorithm incorporating loop closure constraints and magnetic information.
• Utilizing a multi-level Fourier transform for feature extraction from IMU data.
• Employing a RANSAC-based method for validating loop closure matching.
• Applying graph optimization algorithms to suppress cumulative trajectory errors.

Main Results:

• The proposed algorithm provides reliable positioning information without relying on pre-existing maps or extensive geomagnetic data.
• Feature extraction using the multi-level Fourier transform was validated, demonstrating its effectiveness.
• Loop closure detection using RANSAC proved successful, significantly reducing cumulative trajectory errors.
• The trajectory error under loop closure constraint was controlled to an average below 2.15 meters.

Conclusions:

• The foot-mounted IMU positioning algorithm with loop closure constraints offers a robust solution for accurate indoor localization.
• This technique provides a reliable method for collecting precise coordinates for fingerprint database creation.
• The approach demonstrates significant improvements in trajectory accuracy, making it suitable for various indoor positioning applications.