3D Curve Creation on and Around Physical Objects With Mobile AR

Area of Science:

• Computer Vision
• Human-Computer Interaction
• Augmented Reality

Background:

• Mobile Augmented Reality (AR) enables 3D curve creation using smartphones as 3D pens, leveraging motion tracking technologies like Visual Inertial Odometry.
• Tracking robustness issues, such as camera occlusion and textureless environments, hinder precise 3D curve creation in AR, especially during close object interaction.

Purpose of the Study:

• To investigate and minimize tracking errors in mobile AR applications involving 3D curve creation with physical objects.
• To develop a natural interaction system for direct 3D drawing and correction in mobile AR.

Main Methods:

• Conducted an elicitation study focusing on input points and phone grip strategies.
• Performed a quantitative study to analyze tracking errors during mobile AR interactions.
• Developed and evaluated a novel system for direct 3D drawing and interactive error correction in AR.

Main Results:

• Identified key factors influencing tracking accuracy during close interaction between mobile devices and physical objects.
• The proposed system effectively enables direct 3D drawing and interactive correction of 3D curves with reduced tracking errors.
• Demonstrated the system's utility in practical applications like in-situ 3D drawing and direct 3D measurement.

Conclusions:

• The developed system significantly improves the naturalness and accuracy of 3D curve creation in mobile AR.
• This research addresses critical challenges in mobile AR interaction, paving the way for more robust and intuitive AR experiences.
• The findings are valuable for advancing mobile AR applications requiring precise 3D spatial interaction and data capture.