AR Object Recognition Using Database Image Matching
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current augmented reality technologies face limitations in recognizing and interacting with two- or three-dimensional real-world objects, particularly in scenarios where traditional QR codes or hidden signatures are not feasible, leading to restricted recognition areas and the need for new prints or codes for updates, which hinders flexibility and efficiency in providing digital overlays.
Innovation Solution
The system enables automatic recognition of 2D or 3D objects with visible security features, allowing for flexible definition of recognition areas, and updates digital overlays in real-time without altering the original print, using a database to match captured images and provide interactive content, videos, or shopping options, even with older printed materials.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional QR codes or hidden signatures are used for object recognition, then recognition accuracy is improved, but flexibility and adaptability deteriorate because new codes must be printed for updates
Solution Approach 1:
The patent extracts the recognition function from physical QR codes or hidden signatures and relocates it to a database system. The printed material itself becomes the key, and recognition is achieved by capturing images of the material and matching them against stored references in a database, eliminating the need for physical code updates.
Solution Approach 2:
The patent uses digital copies and image processing to create virtual representations of physical objects. By capturing images of printed materials and creating digital matches in a database, the system replaces physical code structures with their digital counterparts, enabling flexible updates without physical changes.
2Adaptability or versatility
If recognition systems are designed to be flexible and adaptable, then adaptability is improved, but recognition precision deteriorates due to restricted recognition areas
Solution Approach 1:
The patent transitions from two-dimensional QR codes to multi-dimensional image recognition. By capturing full images or video of printed materials and analyzing them in a database, the system adds temporal and spatial dimensions to recognition, allowing flexible viewing angles and orientations while maintaining accuracy through comprehensive image matching.
Solution Approach 2:
The patent creates a universal recognition system that can identify various types of printed materials (currency, ID cards, certificates, artworks) using a single database-based approach. The system handles multiple object types and recognition scenarios without requiring specialized codes for each case, achieving both flexibility and precision.
3Reliability
If the system requires new prints or codes for updates, then reliability is improved, but productivity and ease of operation deteriorate
Solution Approach 1:
The patent implements a dynamic update mechanism where the database can be modified without changing the physical printed materials. Updates are achieved by adding new image references or modifying existing ones in the database, allowing real-time or near-real-time updates while maintaining the integrity of the original printed objects.
Solution Approach 2:
The patent introduces a database as an intermediary between the physical printed materials and the recognition system. This intermediary layer allows updates to be made in the digital domain without affecting the physical domain, separating the stability of printed materials from the flexibility of digital data.
4Measurement precision
If the system uses fixed recognition areas, then measurement precision is improved, but adaptability worsens due to restricted viewing angles
Solution Approach 1:
The patent moves from fixed 2D recognition zones to flexible 3D/4D image capture and processing. By accepting images from various angles and orientations and processing them through image recognition algorithms, the system maintains accuracy while accommodating diverse viewing conditions and device orientations.
Data Source
AI summary
Systems, methods and techniques for automatically recognizing two or 3-dimensional real world objects with an augmented reality display device (smartphone or glasses etc.), and augmenting or enhancing the display of such real world objects by superimposing virtual images such as a still or video advertisement, an opportunity to buy, a story or other virtual image presentation. In non-limiting embodiments, the real world object includes visible features including visible security features and a recognition process takes the visible security features into account when recognizing the object and/or displaying superimposed virtual images.