Design and Error Analysis of a Vehicular AR System with Auto-Harmonization

Area of Science:

• Engineering
• Computer Science
• Human-Computer Interaction

Background:

• Augmented reality (AR) systems require high accuracy and low latency for effective use in dynamic environments like vehicles.
• Optical see-through Head-Mounted Displays (HMDs) present unique challenges in maintaining precise alignment and avoiding visual distortions (e.g., "swim").
• Existing AR systems often struggle to meet the stringent requirements for aerospace and ground vehicle applications.

Purpose of the Study:

• To design, develop, and test an AR system tailored for aerospace and ground vehicles.
• To address the challenges of low latency, high tracking accuracy, and precise alignment in optical see-through HMDs.
• To present novel solutions for optical/inertial hybrid tracking systems in AR applications.

Main Methods:

• Development of an optical/inertial hybrid tracking system.
• Implementation of novel solutions for optics, algorithms, synchronization, and alignment.
• Utilizing simulation results to predict registration accuracy.
• Conducting a car test with through-the-eyepiece video capture.
• Performing a detailed covariance analysis of AR registration error.

Main Results:

• Demonstrated a well-registered AR system for vehicles.
• Presented simulation results predicting high tracker accuracy.
• Showcased through-the-eyepiece video evidence of accurate augmentations during driving.
• Derived a detailed covariance analysis of AR registration error.

Conclusions:

• The developed AR system meets stringent accuracy and robustness requirements for vehicles.
• Novel solutions in optical/inertial hybrid tracking effectively address AR challenges.
• The system shows promise for real-world applications in aerospace and ground vehicles.