Real-time Marker-based Tracking of a Non-rigid Object.

Area of Science:

• Computer Vision
• Virtual Reality
• Surgical Simulation

Background:

• Real-time tracking of non-rigid objects is crucial for immersive virtual reality (VR) simulator interfaces.
• Existing feature-based tracking algorithms are insufficient for marker-based, non-rigid object tracking.

Purpose of the Study:

• To present a novel real-time tracking system for non-rigid objects in VR simulators.
• To develop and validate a method for reconstructing 3D positions and updating simulation models based on object deformation.

Main Methods:

• Utilized multiple cameras with integrated image processing to capture marker data (centroid, area, color).
• Developed specialized algorithms for matching camera data to reconstruct 3D marker positions.
• Extracted deformation models from 3D point clouds to update simulation models.

Main Results:

• Achieved real-time tracking of non-rigid objects with adequate latency, robustness, and accuracy.
• Demonstrated the system's effectiveness using a deformable eye interface for the EYESI ophthalmosurgical simulator.
• Validated the approach for tracking deformable tissue in VR applications.

Conclusions:

• The presented marker-based tracking system is suitable for VR simulator applications requiring real-time deformation tracking.
• The methodology is extensible to various simulators involving the tracking of deformable tissues.
• This technology enhances the fidelity and applicability of VR-based training simulators.