Smart Artificial Markers for Accurate Visual Mapping and Localization.

Area of Science:

• Computer Vision
• Robotics
• Artificial Intelligence

Background:

• Artificial marker mapping aids camera localization but requires accuracy improvements for advanced applications.
• Existing methods for Visual Odometry (VO) and Simultaneous Localization and Mapping (SLAM) datasets, robot tracking, and pose estimation can benefit from enhanced accuracy.

Purpose of the Study:

• To improve the accuracy of artificial marker-based map construction and camera localization.
• To introduce a novel 'smart marker' system for more precise mapping and localization.

Main Methods:

• Developed a new mapping method using 'smart markers,' which combine fiducial markers with a pose measurement system (PMS).
• Proposed a localization method that leverages the enhanced accuracy of the smart marker mapping for monocular camera localization with correct scale.
• Collected orientation/distance data from PMS units and calibrated images to estimate marker poses.

Main Results:

• The proposed mapping method reduced Relative Positioning Error (RPE) by 85%.
• The camera localization method decreased Absolute Trajectory Error (ATE) by 50% compared to state-of-the-art techniques.
• Demonstrated significant accuracy improvements in both mapping and localization stages.

Conclusions:

• The smart marker system offers a substantial advancement in artificial marker mapping and camera localization accuracy.
• This enhanced accuracy directly benefits applications requiring precise visual odometry, SLAM, and robotic pose estimation.
• The proposed approach represents a significant improvement over current state-of-the-art methods in the field.