Markerless Joint Center Localization via 3D Point Cloud Projection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for accurately modeling human skeletal structure in immersive environments, such as virtual reality, are either invasive and costly with marker-based systems or inaccurate with markerless systems, necessitating a quick and precise markerless method for estimating limb dimensions and locating centers of rotation in articulated joints.
Innovation Solution
A method involving repetitive movements by the user to acquire 3D positions of bones, computing a center point with minimal standard deviation, transforming the data into a plane, and projecting to find the center of rotation, using hand-held trackers without the need for external sensors or expert assistance.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If marker-based motion capture systems are used to accurately model human skeletal structure, then measurement precision is improved, but device complexity and cost increase
Solution Approach 1:
The patent extracts and removes the complex marker-based sensing hardware from the system, replacing it with a simplified approach that uses only the tracker already present in the virtual reality headset. This eliminates the need for external sensors, motion capture harnesses, and specialized equipment while maintaining the ability to capture sufficient motion data for accurate skeletal modeling.
Solution Approach 2:
The patent creates a virtual copy of the physical joint movement by tracking the position and orientation of the headset tracker throughout the range of motion. Instead of physically marking the body with sensors, the system copies the kinematic data from the single tracker to reconstruct the three-dimensional joint center location through computational geometry algorithms.
2Measurement precision
If marker-based motion capture systems are used to accurately model human skeletal structure, then measurement precision is improved, but setup time and cost increase
Solution Approach 1:
The system performs self-calibration by automatically computing the joint center location from the tracker's own motion data without requiring external measurement tools or expert operators. The user simply performs standardized movements with the headset, and the system autonomously processes the data to determine anatomical landmarks, eliminating the need for time-consuming manual setup procedures.
Solution Approach 2:
The patent incorporates the skeletal modeling calibration into the preliminary setup phase of the virtual reality application itself. By designing the calibration procedure to be integrated into the initial system configuration rather than requiring a separate dedicated setup session, the method ensures accurate skeletal modeling is achieved as part of the normal application startup process.
3Ease of operation
If markerless tracking systems are used to quickly set up virtual reality environments, then ease of operation is improved, but measurement precision deteriorates
Solution Approach 1:
The patent transitions from two-dimensional statistical estimates to three-dimensional geometric computation by utilizing the full spatial trajectory data from the tracker. Instead of relying on population-based averages or simplified 2D projections, the system reconstructs the actual 3D joint center location by analyzing the spatial coordinates and orientations captured throughout the user's range of motion.
Solution Approach 2:
The system changes the fundamental parameter used for joint location determination from statistical body proportions to geometric intersection points. By computing the intersection of spheres or planes defined by the tracker's position at different orientations, the method achieves precise joint localization that adapts to individual anatomy rather than relying on population averages.
Data Source
AI summary
A method for locating a center of rotation of an articulated joint connecting two bones or set of bones of an upper or lower limb of a user, including performing a series of repetitive movements of sweeping one of the bones or set of bones around the joint, and simultaneously acquiring 3D positions of the bone or set of bones during said series, thereby obtaining a 3D cloud of points, computing a point referred to as center point, said center point being a searching point of the 3D searching space for which the standard deviation is the lowest considering the set of distances between the searching point and each point of the 3D cloud of points, transforming the 3D cloud of points into a plane, projecting the center point on said plane, thereby obtaining the center of rotation of the joint.