Cooperative Tracking for Multi-User VR Pose Estimation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional tracking systems for virtual and augmented reality suffer from increased relative error and occlusion issues when multiple users interact closely, leading to degraded performance and misalignment of virtual entities.
Innovation Solution
The system employs a cooperative tracking approach that allows HMDs and hand-held controllers to estimate relative poses between each other, even when globally unavailable due to occlusion, by adjusting measurement frequency based on proximity and using existing components like cameras and LEDs for device-relative tracking.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Area of stationary object
If conventional global tracking systems are used to estimate pose in a mathematical coordinate frame, then tracking coverage area is increased, but relative error between tracked devices increases as users move closer together
Solution Approach 1:
The system segments the tracking approach into two complementary modes: global tracking for general coverage and body-relative tracking for precise relative pose estimation. Each mode operates independently but can be combined, allowing the system to maintain wide coverage while achieving high precision when users are in close proximity.
Solution Approach 2:
The system introduces body-relative tracking as an intermediary mechanism that mediates between global tracking and direct device-to-device measurement. This intermediary approach uses the body as a reference frame to estimate relative pose, bridging the gap when global tracking precision degrades in close-proximity scenarios.
2Measurement precision
If body-relative tracking approaches are used to track hand-held devices with respect to one user, then relative pose accuracy is improved, but the system cannot track devices of multiple users relative to each other
Solution Approach 1:
The body-relative tracking mechanism is designed to be universal and can be applied to track any user's devices with respect to that user's body frame. By making the tracking reference frame user-specific rather than system-specific, the same mechanism works for all users independently, enabling multi-user relative pose estimation.
Solution Approach 2:
The system transitions from a single global coordinate frame to multiple body-relative coordinate frames, adding a dimensional aspect of user-specific reference frames. This allows each user to have their own tracking reference, enabling independent and accurate tracking of multiple users' devices relative to each other.
3Device complexity
If multiple users are tracked independently in open loop fashion, then system complexity is reduced, but virtual entities are not accurately aligned or registered between users
Solution Approach 1:
The system implements feedback by using measured relative poses between users' devices to adjust and align virtual entities across different user views. The relative pose measurements provide feedback information that enables the system to correct misalignments and ensure consistent virtual object positioning for all users.
Solution Approach 2:
The system merges independent open-loop tracking of each user with cooperative relative pose measurement. By combining the simplicity of independent tracking with the alignment information from relative measurements, the system achieves accurate multi-user synchronization without requiring complex fully-coupled tracking infrastructure.
4Ease of operation
If line-of-sight sensing is used for tracking, then measurement simplicity is maintained, but tracking fails when tracking components are occluded by another user or object
Solution Approach 1:
Users' devices serve themselves for tracking by using body-relative measurements that do not require line-of-sight to external sensors. Each user's device can estimate its own pose relative to their body and other users' devices using methods that work independently of external observer visibility, making the system self-sufficient and robust to occlusion.
Data Source
AI summary
This invention relates to tracking of hand-held devices and vehicles with respect to each other, in circumstances where there are two or more users or existent objects interacting in the same share space (co-location). It extends conventional global and body-relative approaches to “cooperatively” estimate the relative poses between all useful combinations of user-worn tracked devices such as HMDs and hand-held controllers worn (or held) by multiple users. Additionally, the invention provides for tracking of vehicles such as cars and unmanned aerial vehicles.