Graph-Based Object Recognition for Similar Shelf Items
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In actual store settings, items of the same category are often placed close together, leading to misrecognition during object recognition processes due to their similar appearances, which existing methods fail to accurately address.
Innovation Solution
An object recognition device that utilizes an image acquisition unit, object recognition unit, graph generation unit, and graph analysis unit to analyze relationships among objects based on their positions, sizes, and features, applying rules such as physical distance, similarity, and positional relationships to enhance recognition accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If individual items are recognized by an object recognition process from a captured image of the item shelf, then the recognition process can identify objects, but misrecognition occurs due to similar appearances of items placed close to each other
Solution Approach 1:
The patent transitions from analyzing only individual item images to constructing a graph structure that captures spatial relationships and contextual information between multiple items. This dimensional expansion from 2D image analysis to graph-based relational analysis enables the system to distinguish between similarly appearing items by considering their positional relationships and surrounding context.
Solution Approach 2:
The patent introduces a graph structure as an intermediary representation between the captured image and the final recognition result. This graph acts as a mediator that encodes spatial relationships, item categories, and contextual information, allowing the system to resolve ambiguities in individual item recognition through relational reasoning.
2Productivity
If items of the same category are placed close to each other on the item shelf, then storage efficiency is improved, but misrecognition occurs during object recognition processes
Solution Approach 1:
The patent merges individual item recognition with spatial relationship analysis by constructing a unified graph structure that incorporates both item attributes and their positional relationships. This combined approach allows the system to leverage contextual information from neighboring items to improve recognition accuracy while maintaining efficient shelf utilization.
Solution Approach 2:
The patent implements a feedback mechanism where the graph analysis results are used to refine and correct individual item recognition. The spatial relationships and contextual information derived from the graph structure provide feedback that helps disambiguate similarly appearing items, improving overall recognition accuracy without requiring changes to physical item placement.
Data Source
AI summary
In an object recognition device, an image acquisition means acquires an image. A recognition means recognizes each of objects included in an image, and generates each recognition result. A graph generation means defines each of the recognized objects on one of a node and an edge and defines each of relationships among the objects on another one of the node and the edge, based on each recognition result. A graph analysis means analysis a graph and generates an analysis result indicating the relationships among the objects.