Object Recognition Using Visual Identifiers to Reduce Storage
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing object recognition systems face challenges in accurately identifying objects due to variations in lighting conditions, angles, and distances, leading to the need for large template databases and excessive computational resources, especially when using appearance-based methods, and slow processing with feature-based methods.
Innovation Solution
The implementation of a visual identifier, such as a QR code or text-based identifier, on objects allows object recognition systems to generate more accurate reference models by capturing and processing these identifiers to obtain classifier information from external databases, reducing the need for extensive template generation and computational resources.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If appearance-based methods are used for object recognition, then recognition speed is improved, but storage requirements and computational resources increase significantly due to the need for large template databases
Solution Approach 1:
The patent extracts only the essential visual identifier features from objects rather than storing complete appearance templates. By capturing key distinguishing features through visual identifiers and storing only these extracted characteristics in the database, the system maintains fast recognition capability while dramatically reducing storage requirements compared to storing full appearance-based templates.
2Use of energy by moving object
If feature-based methods are used for object recognition, then computational resources are reduced, but processing speed decreases
Solution Approach 1:
The patent applies preliminary action by pre-extracting and storing visual identifier features in the database before recognition is needed. When an object needs to be recognized, the system retrieves these pre-processed visual identifiers and compares them directly with captured images, eliminating the need for complex real-time feature extraction and significantly improving processing speed while maintaining low computational resource usage.
3Measurement precision
If visual identifiers are used to generate reference models, then object recognition accuracy is improved, but the complexity of the system increases due to additional processing steps
Solution Approach 1:
The patent introduces visual identifiers as an intermediary element between the object and the recognition system. These visual identifiers serve as mediators that encode object information in a standardized, easily comparable format. By using this intermediary layer, the system achieves high recognition accuracy without requiring complex processing algorithms, as the visual identifiers provide clear distinguishing features for straightforward comparison and matching.
Data Source
AI summary
Methods, apparatus, systems, and articles of manufacture are disclosed to improve robot object recognition. An example apparatus includes a visual object recognizer to obtain a visual identifier associated with a target object, and a recognizable object model generator to generate a model of the target object based on mapping an image of the target object to classifier information corresponding to the visual identifier.