Real-Time Motion Tracking for Mobile Augmented/Virtual Reality Using Adaptive Visual-Inertial Fusion.

Area of Science:

• Computer Vision
• Robotics
• Human-Computer Interaction

Background:

• Real-time 6-Degree of Freedom (DoF) motion tracking is critical for mobile augmented/virtual reality (AR/VR) to align virtual elements with the real world.
• Mobile device limitations cause latency in pose estimation, degrading AR/VR user experience.

Purpose of the Study:

• To develop a real-time visual-inertial motion tracking system for mobile AR/VR applications.
• To address latency and jitter issues in mobile AR/VR motion tracking.
• To enhance the robustness of motion tracking, especially under motion blur conditions.

Main Methods:

• Utilizes high-frequency inertial sensor data to achieve real-time pose estimation.
• Implements an adaptive filter framework to manage visual-inertial fusion, mitigating jitter and latency.
• Enhances traditional visual-only tracking for improved performance during motion blur.

Main Results:

• The proposed visual-inertial system achieves real-time 6-DoF motion tracking.
• The adaptive filter effectively balances jitter and latency for improved user experience.
• Enhanced visual-only tracking improves AR/VR performance in the presence of motion blur.

Conclusions:

• The developed visual-inertial motion tracking method provides smooth and robust 6-DoF tracking for mobile AR/VR.
• This approach significantly improves the user experience in mobile AR/VR by reducing tracking latency and jitter.
• The system demonstrates practical applicability for real-time mobile AR/VR applications.