Virtual Content Pose Determination Using Stored Anchor Data
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing mixed reality and virtual reality technologies face high computational intensity and battery drain due to the need for detailed image analysis and visual feature matching to determine the pose of virtual content, especially when re-entering previously visited locations.
Innovation Solution
A method that anchors the content pose relative to an external device, allowing for efficient determination of the content pose by retrieving stored pose data when re-entering a space, reducing the need for resource-intensive visual feature matching by using a pose uncertainty indicator and movement flags to decide between stored and visual feature-based pose determination.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If visual search and matching is performed to determine content pose by inspecting environment and performing detailed image analysis, then content pose can be accurately determined, but computational intensity and battery drain increase significantly
Solution Approach 1:
The system performs preliminary action by storing visual features and their corresponding poses during initial environment inspection. When determining content pose later, the system retrieves pre-stored visual features and performs lightweight matching instead of repeating full image analysis, significantly reducing computational intensity and battery drain while maintaining pose determination accuracy.
2Measurement precision
If detailed image analysis is performed to infer specific visual features and perform matching, then content pose can be determined, but processing power requirements increase significantly
Solution Approach 1:
The system extracts and stores visual features during an initial phase when processing power is more readily available. During subsequent pose determination operations, the system performs lightweight matching against pre-extracted features rather than performing full image analysis, dramatically reducing processing power requirements while preserving matching accuracy.
Solution Approach 2:
The system extracts essential visual features from full images and stores only the extracted features for future matching. This extraction process separates the computationally intensive feature extraction step from the lightweight matching step, allowing the system to perform accurate pose determination with minimal processing power during actual use.
3Measurement precision
If visual feature matching is performed continuously to determine content pose, then pose accuracy is maintained, but time consumption increases
Solution Approach 1:
The system performs preliminary extraction and storage of visual features from the environment during an initial phase. When pose determination is needed, the system performs rapid matching against pre-stored features instead of repeating full image analysis, reducing pose determination time from potentially seconds to milliseconds while maintaining accuracy.
Data Source
Figure 1~2B
Figure 3A
Figure 3B
AI summary
It is provided a method for determining a content pose of a piece of virtual content. The method is performed in a pose determiner and comprises the steps of: acquiring a first anchor pose of the anchoring device; determining a first content pose, being a pose of a piece of virtual content, in relation to the first anchor pose; storing the first content pose in relation to the first anchor pose; determining that a user device needs to determine a content pose of the piece of virtual content; acquiring a second anchor pose of the anchoring device; retrieving the stored first content pose; determining a second content pose of the piece of virtual content based on the first content pose and the second anchor pose; and rendering the piece of virtual content in the second content pose on a user device comprising a display.