An Enhancement of Outdoor Location-Based Augmented Reality Anchor Precision through VSLAM and Google Street View

Area of Science:

• Computer Vision
• Geographic Information Systems
• Augmented Reality

Background:

• Outdoor Location-Based Augmented Reality (LAR) demands high positioning accuracy for integrating virtual content.
• Traditional methods using Global Positioning System (GPS) and compasses often lack the precision required for accurate AR alignments.
• Existing sensor fusion techniques in LAR applications face limitations in achieving precise real-world AR content placement.

Purpose of the Study:

• To introduce and evaluate an innovative approach for enhancing the precision of LAR anchors in outdoor settings.
• To address the accuracy limitations of conventional sensor fusion methods in outdoor LAR applications.
• To improve the alignment of augmented reality content with real-world environments.

Main Methods:

• Leveraged Visual Simultaneous Localization and Mapping (VSLAM) technology for enhanced localization.
• Utilized cloud-based methodologies combined with the Google Street View (GSV) visual reference database.
• Conducted experiments using 10 Points of Interest (POI) as anchor point coordinates for evaluation.

Main Results:

• Demonstrated substantial enhancements in overall positioning accuracies compared to conventional GPS-based approaches.
• Achieved superior accuracy in aligning AR anchor content in the real world.
• Performance testing confirmed improved accuracy and load handling for the proposed LAR method.

Conclusions:

• The integration of VSLAM and GSV offers a significant improvement in outdoor LAR anchor precision.
• This innovative approach overcomes the accuracy limitations of traditional GPS and sensor fusion methods.
• The findings pave the way for more immersive and accurate outdoor augmented reality experiences.