Head-Display Pose Estimation With Motion-Capture Calibration

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing head display devices in extended reality (XR) systems face inaccuracies in pose estimation due to errors in determining camera movement trajectories, leading to suboptimal adjustment of pose estimation parameters and reduced display effectiveness.

Innovation Solution

A method and apparatus that utilize a data processing system to determine a first transformation matrix from an object coordinate system to a camera coordinate system, adjusting pose estimation parameters based on the actual movement trajectory of a motion capture target object, thereby improving the accuracy of pose estimation and display performance.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If hand-eye calibration is used to adjust pose estimation parameters based on camera movement trajectory, then pose estimation accuracy can be improved, but trajectory errors are amplified causing larger evaluation errors

Engineering Contradiction:
Improvepose estimation accuracyVSAvoidevaluation result reliability
Core Design Contradiction:
Measurement precisionVSReliability

Solution Approach 1:

The patent introduces a motion capture target object as an intermediary between the camera and the hand-eye calibration process. Instead of directly calibrating based on camera trajectory alone, the system uses the motion capture target object (with known rigid connection to the head display device) as a mediator to establish a more reliable transformation relationship. This intermediary provides additional constraint information that breaks the error amplification chain in traditional hand-eye calibration.

Inventive Principle:
Principle #24Intermediary (Mediator)

Solution Approach 2:

The system implements a feedback mechanism where the motion capture system continuously tracks the motion capture target object and provides real-time position and orientation data. This feedback is used to continuously refine and adjust the pose estimation parameters, creating a closed-loop system that compensates for errors dynamically rather than relying on a single calibration step that amplifies trajectory errors.

Inventive Principle:
Principle #23Feedback

2Ease of operation

If camera movement trajectory is determined through hand-eye calibration, then pose estimation can be adjusted, but trajectory errors cause larger evaluation errors

Engineering Contradiction:
Improvepose estimation adjustmentVSAvoidtrajectory evaluation precision
Core Design Contradiction:
Ease of operationVSMeasurement precision

Solution Approach 1:

The patent merges multiple data sources and calibration approaches into a unified system. It combines hand-eye calibration with motion capture technology, integrating camera coordinate system transformations with motion capture target object tracking. This merging creates a more robust system where the strengths of each method compensate for the weaknesses of the other, particularly reducing the impact of trajectory errors on evaluation precision.

Inventive Principle:
Principle #5Merging (Combining)

Data Source

PatentEP4614443A1Data processing method and apparatus, device, and storage medium
Publication Date: 2025.09.10 TENCENT TECHNOLOGY (SHENZHEN) CO LTD
  • EP4614443A1 patent drawingFigure 1
  • EP4614443A1 patent drawingFigure 2
  • EP4614443A1 patent drawingFigure 3a~3d

AI summary

A data processing method and apparatus, a device, and a storage medium are provided. The method includes: obtaining at least one calibration target image captured by a camera of a head display device during a movement of the head display device within a motion capture area; determining a first transformation matrix from an object coordinate system, in which a motion capture target object is located, to a camera coordinate system corresponding to a reference camera, the reference camera being any camera of the head display device, and the motion capture target object being rigidly connected to the head display device; and adjusting, based on the first transformation matrix and a first movement trajectory of the motion capture target object in a world coordinate system when the head display device moves within a preset time period, a pose estimation parameter of the head display device.