Similarity-Based Order Management for Merchandise Recognition
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image recognition systems in retail stores suffer from high erroneous recognition rates of merchandise items, particularly when multiple items are combined, leading to potential sales losses and increased manual intervention.
Innovation Solution
An order management apparatus and method that calculates similarity degrees between merchandise items to determine erroneous recognition risks, allowing for user intervention or automatic adjustments to reduce these risks through merchandise item changes or optimization learning based on background image data.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If image recognition is used for checkout to automate the process, then productivity is improved, but erroneous recognition rate increases
Solution Approach 1:
The system performs preliminary actions by calculating similarity degrees between merchandise items and determining erroneous recognition risks before the actual checkout process. The order management apparatus proactively identifies potential recognition errors by analyzing image data of multiple merchandise items and computing similarity metrics, allowing for preventive measures to be taken before erroneous recognition occurs during checkout.
2Productivity
If multiple merchandise items are processed together to handle combinations, then productivity is improved, but erroneous recognition rate increases
Solution Approach 1:
The system calculates similarity degrees between multiple merchandise items in advance and determines erroneous recognition risks for item combinations before checkout. The order management apparatus analyzes image data of multiple items simultaneously, computes similarity metrics, and identifies potential recognition errors proactively, enabling the system to handle multiple items efficiently while maintaining accuracy through preventive risk assessment.
3Reliability
If manual intervention is increased to verify recognition results, then reliability is improved, but productivity decreases
Solution Approach 1:
The system implements feedback mechanisms by providing information on erroneous recognition risks to the order management apparatus, which can then alert operators or adjust processing based on the calculated risks. The control apparatus uses feedback from similarity degree calculations and risk assessments to determine when manual verification is necessary, allowing for intelligent allocation of manual intervention to high-risk cases while maintaining high-speed processing for low-risk items.
Data Source
AI summary
An order management device is provided with: an input circuit that acquires information about a plurality of articles including an order candidate article and to be subjected to image recognition after the order candidate article is ordered; a determination circuit that determines, with respect to two or more different articles included in the plurality of articles, an index indicating a possibility that the articles are incorrectly recognized as the same article when the image recognition is performed, on the basis of information about the similarity among the plurality of articles in the image recognition; and an output circuit that outputs information about the determined index.