AR Poster Recognition Without Markers
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current augmented reality technologies face limitations in recognizing and interacting with real-world objects, particularly 2D and 3D printed items, as they often require specific markers or restricted areas of recognition, which can restrict flexibility and usability, especially in varying lighting conditions and angles.
Innovation Solution
The system employs image recognition techniques that can automatically identify 2D and 3D objects without the need for QR codes or hidden signatures, using flexible areas of recognition and machine learning to trigger digital overlays, enabling the superimposition of virtual content such as videos, information, and shopping options directly onto recognized objects, even in low-light conditions and from different angles.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional AR systems use QR codes or hidden signatures for object recognition, then recognition accuracy is improved, but flexibility and usability are reduced
Solution Approach 1:
The patent removes QR codes and hidden signatures from the recognition process, extracting only the essential visual features of objects. The system captures images of objects and uses machine learning to identify them without requiring any artificial markers, thereby improving flexibility while maintaining recognition accuracy through advanced image processing algorithms.
Solution Approach 2:
The system creates a universal recognition mechanism that can identify any object within the field of view without requiring object-specific markers. The machine learning model is trained to recognize diverse objects (products, posters, packaging) using only their natural visual appearance, making the system adaptable to various applications without needing to modify the recognition approach.
2Reliability
If AR systems require specific markers for recognition, then recognition reliability is improved, but device complexity increases
Solution Approach 1:
The patent replaces the mechanical approach of placing physical markers (QR codes, hidden signatures) on objects with a computational approach using machine learning. The system uses neural networks and image processing algorithms to automatically identify objects based on their visual features, eliminating the need for manual marker placement and reducing overall system complexity.
Solution Approach 2:
The system performs self-training and self-improvement through machine learning. The neural network automatically learns to recognize objects from captured images without requiring pre-programmed marker recognition logic. The system adapts to different objects and conditions autonomously, reducing the need for manual configuration and complex recognition rules.
3Speed
If AR recognition uses restricted areas, then processing speed is improved, but adaptability to different conditions deteriorates
Solution Approach 1:
The patent implements dynamic image processing that adapts to varying lighting conditions and viewing angles. The machine learning model processes the entire captured image dynamically, adjusting recognition parameters based on environmental conditions. This allows the system to maintain both processing speed and adaptability by using efficient algorithms that can handle variable inputs without requiring restricted recognition areas.
Data Source
AI summary
A poster information system recognizes a real world object relating to a poster based transaction and uses augmented, mixed and/or virtual reality to display information relating to the transaction on a smart device such as a smartphone or AR glasses.