Head-Mounted Display Pose Tracking with Hand-Based Drift Calibration
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing self-tracking devices in head-mounted displays suffer from accuracy issues due to error accumulation over time, leading to incomplete or inaccurate six-degree-of-freedom information, which affects user experience and interactive operations.
Innovation Solution
A head-mounted display system that calculates drift information between hand pose and self-tracking pose to calibrate self-tracking pose information, generating six-degree-of-freedom data even when self-tracking devices provide incomplete data, and fuses pose information for continuous display accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Duration of action of stationary object
If self-tracking device operates over time, then tracking function is maintained, but error values accumulate reducing accuracy
Solution Approach 1:
The system continuously compares hand pose information (six-degree-of-freedom) with self-tracking pose information to calculate drift values, then uses this feedback to calibrate and correct the self-tracking device's pose information, maintaining accuracy over extended operation periods
Solution Approach 2:
The head-mounted display system performs self-calibration by generating its own reference data (hand pose information) and using it to correct its self-tracking device's accumulated errors, enabling autonomous drift compensation without external intervention
2Reliability
If self-tracking device enters three-degree-of-freedom mode, then device can continue operation during blocking or relocation, but complete six-degree-of-freedom information cannot be provided
Solution Approach 1:
The system uses hand pose information as an intermediary reference that provides complete six-degree-of-freedom data during periods when self-tracking device cannot provide full information, enabling continuous and accurate pose tracking through data fusion and calibration
Solution Approach 2:
The system dynamically changes the parameter completeness by switching between using self-tracking device data and hand pose information based on data quality, maintaining six-degree-of-freedom information availability even when self-tracking device operates in limited three-degree-of-freedom mode
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A head-mounted display, pose tracking system, and method are provided. The head-mounted display generates a plurality of first real-time images including a user holding a self-tracking device in a physical space. The head-mounted display receives first self-tracking pose information from the self-tracking device. The head-mounted display generates a hand pose information based on the first real-time images, and the hand pose information is a six-degree-of-freedom information. The head mounted display calculates drift information of the hand pose information and the first self-tracking pose information. The head mounted display calibrates the first self-tracking pose information based on the drift information.