Multi-sensor calibration of low-cost magnetic, angular rate and gravity systems

Area of Science:

• Biomedical Engineering
• Sensor Technology
• Human Motion Analysis

Background:

• Low-cost nine degrees-of-freedom (9DOF) magnetic, angular rate, and gravity (MARG) sensor systems are increasingly used in body sensor networks (BSNs).
• Accurate calibration is crucial for reliable data from MARG sensors, especially in systems like the Integrated Posture and Activity Network by Medit Aachen (IPANEMA).
• Existing calibration methods may have limitations in integration and accuracy for complex motion analysis.

Purpose of the Study:

• To present a novel, efficient calibration procedure for 9DOF MARG sensor systems.
• To enhance the accuracy and reliability of BSNs for posture and activity monitoring.
• To enable precise orientation alignment between MARG sensors and external motion capture systems.

Main Methods:

• Development of a calibration procedure utilizing a calibration cube, reference table, and a body sensor network (BSN).
• Integration of two sensor nodes within the IPANEMA BSN for calibration and reference measurements.
• Implementation of a novel algorithm processing arbitrarily-executed motions to refine calibration.
• A two-stage experimental study involving offset minimization and dynamic deviation correction.

Main Results:

• The new calibration procedure demonstrated effective integration of sensor nodes using a calibration cube.
• The algorithm successfully processed arbitrary motions, improving calibration performance.
• The routine facilitated alignment with motion capture inertial reference systems, minimizing orientation errors.
• Experimental validation showed promising results with a maximum Root Mean Square (RMS) error of 3.89°.

Conclusions:

• The proposed calibration procedure offers an effective solution for low-cost 9DOF MARG sensor systems.
• This method enhances the accuracy of BSNs, particularly the IPANEMA system, for various applications.
• The approach provides a reliable way to minimize errors in orientation and improve movement model accuracy.