Related Experiment Video
Updated: Aug 5, 2026

Home-Based Monitor for Gait and Activity Analysis
Published on: August 8, 2019
MVO: A Magneto-Visual Odometry System for Indoor Positioning
Tongxing Peng1, Chao Ming1, Zhengpeng Yang1
1College of Mechanical Engineering, Nanjing University of Science and Technology, Nanjing 210094, China.
This study introduces a magneto-visual odometry (MVO) system for robust indoor robot navigation. By combining magnetic field data with visual odometry, MVO enhances positioning accuracy in challenging environments.
Area of Science:
- Robotics
- Computer Vision
- Sensor Fusion
Background:
- High-precision indoor positioning is crucial for mobile robots in GPS-denied areas.
- Visual odometry struggles with illumination changes and textureless environments.
- Existing methods lack resilience in challenging indoor conditions.
Purpose of the Study:
- To develop a magneto-visual odometry (MVO) framework for enhanced indoor robot localization.
- To leverage indoor magnetic field anomalies as complementary constraints for visual odometry.
- To improve the robustness and accuracy of mobile robot navigation in GNSS-denied environments.
Main Methods:
- Integration of a 30-magnetometer planar array with a stereo camera for multi-modal perception.
- Utilizing magnetic field gradient information in the frontend for relative-pose constraints and feature matching.
- Employing factor graph optimization (FGO) with iSAM2 in the backend to fuse magnetic and visual factors.
- Simulation-based analysis of magnetometer configuration, sensor count, and calibration sensitivity.
Main Results:
- MVO demonstrated improved localization accuracy compared to MSCKF-Stereo and VINS-Fusion baselines in synthesized magnetic field data.
- The framework showed a moderate computational load.
- Frontend simulations validated the impact of magnetometer configuration and calibration on magnetic relative-pose estimation.
- Trajectory-level evaluations confirmed the feasibility of magnetic-visual fusion for indoor odometry.
Conclusions:
- The proposed MVO framework offers a viable approach to enhance indoor robot positioning by integrating magnetic field data with visual odometry.
- Simulation results validate the potential of magnetic constraints for improving tracking continuity and accuracy under visual degradation.
- Further research is needed for real-world validation with actual magnetometer array measurements.
Related Concept Videos
Design Example: Identifying the Locations of Monuments in the Field Using Global Positioning System Device
Electronic Distance Measuring Instruments
Galvanometer
The galvanometer consists of two concave-shaped permanent magnets, providing a uniform radial magnetic field in the annular region. In the center, a pivoted coil of fine copper wire is placed in the uniform magnetic...
Field Application of Global Positioning System
Gyroscope
Magnetic Vector Potential
Consider an ideal solenoid with n turns per unit length and radius R. If I is the current through the solenoid, the magnetic field inside the solenoid is expressed as the product of vacuum...