AR Object Recognition with Stored Geometry for Content Placement
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current augmented reality (AR) systems face challenges in creating AR content due to the requirement of technical skills in graphic design or computer-aided design (CAD), limiting content adoption for mainstream users.
Innovation Solution
An AR system utilizing visual object recognition and stored geometry to create and render virtual objects, allowing users to annotate objects with AR content using a single device, such as a phone or AR headset, by detecting the object's three-dimensional geometry and mapping it within the scene, enabling the placement of virtual objects at fixed orientations relative to the object, which remain consistent regardless of the device's location.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If traditional AR content creation methods are used, then AR content can be created with high precision, but technical skills in graphic design or CAD are required
Solution Approach 1:
The system creates a digital copy of the real-world object by capturing its image and generating a three-dimensional geometry model. This copy serves as the basis for AR content creation, eliminating the need for manual modeling while preserving the object's accurate geometry and spatial relationships
Solution Approach 2:
The patent replaces manual graphic design and CAD operations with automated computer vision and machine learning algorithms. The system automatically detects objects, extracts their three-dimensional geometry, and enables AR content placement without requiring users to interact with complex design tools
2Productivity
If manual AR content creation is used, then content can be created with precise control, but the process is time-consuming
Solution Approach 1:
The system performs preliminary actions by automatically capturing object images and generating three-dimensional geometry models before AR content creation begins. This pre-processing eliminates the time-consuming manual modeling step and prepares the data structure needed for rapid content placement
Solution Approach 2:
The system enables self-service AR content creation by allowing users to simply capture an object with their device camera. The system then automatically handles object detection, geometry extraction, and content placement, eliminating the need for users to invest significant time in manual content creation processes
3Manufacturing precision
If AR content is anchored to real-world objects, then content placement accuracy is improved, but the system complexity increases
Solution Approach 1:
The patent introduces an intermediary three-dimensional geometry model that mediates between the real-world object and the AR content. This model serves as a reference framework that enables accurate content placement by providing spatial relationships and geometric constraints without requiring direct complex interactions between the device and the physical object
Data Source
AI summary
A three-dimensional geometry of an object is stored, and a representation of the object that can facilitate identification of the object via a camera is also stored. An image of the object is obtained via a scene in the camera. The presence of the object is detected from the image via the stored representation of the object. The three-dimensional geometry to the object is mapped within the scene based on the detection. A virtual object is attached to a point that is at a fixed orientation relative to the three-dimensional geometry based on the mapping. The virtual object is rendered on an augmented reality display as being located at the point and at the fixed orientation regardless of a change in location of the augmented reality display relative to the object.