MoiréBoard: A Stable, Accurate and Low-cost Camera Tracking Method

Force Feedback & Pseudo-Haptic Weight3D Modeling & AnimationSoftware Engineers & DevelopersHCI Researchers

Title of the Paper

MoiréBoard: A Stable, Accurate, and Low-cost Camera Tracking Method

Paper Information

  • Subject Area: Human-Computer Interaction and Computer Vision
  • Keywords: MoiréBoard, Visual Localization, Camera Tracking, Virtual Reality (VR), Augmented Reality (AR), Marker Tracking, Low Cost, 3D Printing, High Precision, Optical Signals

Research Background and Problem

  • Problem or Challenge:

    • Commercial VR and AR systems often rely on expensive specialized hardware (such as base stations emitting lasers) for device tracking. While these systems provide extremely high localization accuracy, their high cost makes them difficult to adopt in low-budget scenarios.
    • Low-cost solutions such as inertial measurement units (IMU) and visual markers can perform device localization but suffer from significant tracking errors, especially in positional tracking, making it challenging to meet high-precision requirements.
  • Importance of the Research:

    • Spatial device localization is a critical foundation for VR and AR applications. Addressing the dilemma between high cost and low accuracy can help promote the widespread adoption and application of virtual reality technology.
  • Motivation and Related Work:

    • Current optical tracking methods include active marker tracking based on external base stations (e.g., Lighthouse systems) and systems based on SLAM algorithms. However, these methods significantly increase costs or exhibit instability in certain visual environments.
    • The authors aim to challenge the trade-off between cost and performance by proposing a low-cost tracking method based on optical visual markers that is both efficient and highly accurate.

Solution

  • Proposed Method or Solution:

    • A novel optical visual marker, MoiréBoard, is introduced, utilizing the visual phenomenon of moiré patterns to track camera position information. This method does not require camera intrinsic calibration or additional power supply equipment.
  • Innovative Aspects of the Solution:

    • Exploits the moiré effect to amplify device displacement, thereby improving localization accuracy.
    • Proposes formulas based on marker physical parameters and dimensionless ratios to eliminate dependence on camera calibration.
    • Constructs an efficient tracking algorithm by systematically analyzing the formation process of moiré patterns and the camera projection relationship, eliminating the need for calibration.
  • Implementation Steps and Key Techniques:

    1. Create the MoiréBoard using 3D printing to construct the front-layer grid, with the back layer consisting of either an LED display or printed paper.
    2. Use a camera to capture moiré patterns in real-time and combine visual marker localization to identify the pattern region.
    3. Calculate the camera's position information based on image processing and mathematical models.
    4. Validate the tracking algorithm's accuracy and advantages over peers through experiments.

Research Results

  • Specific Results:

    • Achieved a full-field 3-DOF camera tracking method without requiring camera calibration.
    • Experimental results show that MoiréBoard's localization accuracy is comparable to high-end VR tracking devices like Lighthouse 2.0, while costing significantly less than commercial systems.
    • Developed a precision analysis method based on the moiré effect, achieving sub-pixel level error in tests.
  • Comparative Advantages Over Existing Solutions:

    • Compared to traditional chessboard marker methods, MoiréBoard improves localization accuracy by an order of magnitude.
    • Does not rely on specialized hardware, is simple to produce, and is low-cost, making it suitable for mobile devices and low-budget projects.
    • Demonstrates stronger robustness against low-quality images and varying video frame rates.
  • Experimental or Evaluation Results:

    • Synthetic scene experiments show stable high-precision tracking capabilities across different positions and rotation angles.
    • Real-world tests indicate that MoiréBoard achieves millimeter-level error within a range of 1–2.5 meters, comparable to the precision of commercial Lighthouse 2.0 systems.
    • Performance validation in domain applications (e.g., Google Cardboard platform) demonstrates its ability to effectively support real-time 6-DOF VR experiences.
  • Limitations and Future Directions:

    • Limited by the camera's field of view (FoV), motion blur may occur during high-speed movements.
    • The current design requires four additional visual markers for image correction; future research could explore direct correction methods based on moiré patterns to further optimize the board design.
    • The working range of a single MoiréBoard is relatively limited, but collaborative use of multiple boards could expand range and precision while enabling device orientation tracking.

Conclusion

By leveraging the amplification characteristics of the moiré pattern effect, this paper proposes a stable, efficient, and low-cost tracking method for robots and VR/AR devices. The method overcomes the precision bottleneck of traditional visual markers and achieves high robustness across a wide range. Future work could focus on optimizing board design, expanding application scenarios, and exploring integration with other signals or algorithms to enhance overall performance.

Quick Actions

Share

Share this page

ios_share

https://hci.top/en/papers/uist/61403/2021

AdRecommended

Learn AI Coding at CodeNow

open_in_newOpen DOI Link
DOI: https://doi.org/10.1145/3472749.3474793
At a Glance

Paper Snapshot

fact_check
dataset
Source
UIST
calendar_month
Year
2021
emoji_events
Award
No award tagged
group
Authors
2 authors
sell
Subtopics
Force Feedback & Pseudo-Haptic Weight, 3D Modeling & Animation
work
Professions
Software Engineers & Developers, HCI Researchers
article
Content Status
Full text indexed
hub
Related Papers
3 related papers