3D Model Generation Without Markers Using Pose Estimation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for generating three-dimensional model data of objects fail when marker information cannot be recognized, hindering accurate augmentation of virtual objects in augmented reality.
Innovation Solution
A method involving a computing device that captures images of objects with and without markers, generates view data, and estimates relative pose data to create three-dimensional model data, even in situations where marker information is unavailable, by using camera view data and pose data to calculate object pose and orientation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If marker information is used to generate three-dimensional model data, then measurement precision is improved, but reliability deteriorates when marker information cannot be recognized
Solution Approach 1:
The patent introduces an intermediary estimation process that uses camera view data and previously obtained pose data as mediators to bridge the gap when marker information is unavailable. Instead of directly relying on marker recognition, the system uses these intermediary data sources to estimate relative pose data, maintaining system reliability during marker recognition failures.
Solution Approach 2:
The patent stores camera view data and pose data in advance before marker recognition failures occur. By performing preliminary data collection and storage during normal operation, the system prepares alternative data sources that can be utilized when marker information becomes unavailable, ensuring continuous three-dimensional model data generation.
2Manufacturing precision
If marker information is required for generating three-dimensional model data, then manufacturing precision is improved, but adaptability deteriorates in situations where markers are not visible
Solution Approach 1:
The patent implements a dynamic data generation approach that adapts between two modes: using marker information when available for high precision, and switching to estimation based on camera view data and stored pose data when markers are not visible. This dynamic switching mechanism maintains manufacturing precision while significantly improving adaptability to different operational conditions.
Solution Approach 2:
The patent changes the input parameters used for three-dimensional model data generation based on marker availability. When markers are visible, it uses marker-derived relative pose data; when markers are not visible, it transitions to using camera view data and stored pose data as alternative parameters, thereby maintaining adaptability across varying environmental conditions.
3Measurement precision
If only marker-based methods are used, then measurement accuracy is improved, but productivity deteriorates due to complete failure when markers cannot be recognized
Solution Approach 1:
The patent performs preliminary storage of camera view data and pose data during normal operation when markers are visible. This advance preparation ensures that when marker recognition fails, the system can immediately switch to using stored data without interruption, maintaining productivity while preserving measurement accuracy through the alternative estimation pathway.
Solution Approach 2:
The patent ensures continuous generation of three-dimensional model data by maintaining multiple data sources and switching between them based on marker availability. The useful action of data generation continues uninterrupted by seamlessly transitioning from marker-based methods to estimation-based methods, preventing complete failure and maintaining productivity.
Data Source
AI summary
The method of generating three-dimensional model data of an object includes estimating pose data of the object based on the first view data of the camera and first relative pose data of the object, and estimating second relative pose data of the object based on the pose data of the object and the second view data of the camera.