Head-Mounted Rendering Compensation for Posture Delay Correction

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing head-mounted devices suffer from poor accuracy in rendering compensation due to the lack of consideration for user posture changes during network transmission and image rendering delays.

Innovation Solution

A control method for head-mounted devices that includes sending posture data to a cloud server, recording real posture change data, and performing rendering compensation based on this data to obtain accurate rendered images, which involves calculating rotation angles and adjusting image content ranges to account for delays and user movements.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If asynchronous time warp compensation is performed based on historical posture data, then rendering compensation can be executed, but the accuracy deteriorates due to not considering user posture changes during network transmission and rendering time

Engineering Contradiction:
Improverendering compensation accuracyVSAvoidnetwork transmission time and rendering time
Core Design Contradiction:
Measurement precisionVSLoss of time

Solution Approach 1:

The patent applies preliminary action by recording posture change data during the network transmission and rendering period before the compensation is needed. The head-mounted device continuously captures and stores posture data from the first target time (when rendering request is sent) to the second target time (when rendered image is received), so that the accurate compensation can be performed using this pre-recorded data without waiting for additional measurements after the delay period.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The patent implements feedback by using the recorded posture change data (from the time when rendering request is sent to when rendered image is received) to correct and adjust the rendering compensation. This feedback loop ensures that the compensation accurately reflects the user's actual posture changes during the delay period, improving the overall accuracy of the rendering system.

Inventive Principle:
Principle #23Feedback

2Manufacturing precision

If rendering compensation is performed without considering posture changes during delay period, then processing can be simplified, but the rendering quality deteriorates due to inaccurate compensation

Engineering Contradiction:
Improverendering qualityVSAvoidposture data recording and processing complexity
Core Design Contradiction:
Manufacturing precisionVSDevice complexity

Solution Approach 1:

The patent applies continuity of useful action by continuously recording posture change data from the first target time to the second target time without interruption. This continuous capture ensures that all posture changes during the network transmission and rendering delay period are captured, maintaining rendering quality while the recording process runs parallel to other system operations without adding significant complexity.

Inventive Principle:
Principle #20Continuity of useful action

Data Source

PatentUS12493186B2Control method for head-mounted device and image rendering method
Publication Date: 2025.12.09 GOERTEK INC
  • US12493186B2 patent drawing
  • US12493186B2 patent drawing
  • US12493186B2 patent drawing

AI summary

Disclosed are a control method and apparatus for a head-mounted device, and a head-mounted device, as well as an image rendering method and apparatus, and an electronic device. The control method comprises: sending a rendering request containing posture data to a cloud server, and receiving a first rendered image, wherein the first rendered image is an image obtained by rendering in the cloud server based on the posture data; obtaining real posture change data from a first target time to a second target time, wherein the first target time is a time when the rendering request is sent, and the second target time is a time when the first rendered image is received; and performing rendering compensation on the first rendered image based on the posture change data to obtain a second rendered image.