Object Recognition Device for Checkout Accuracy
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In checkout systems, improper collation data in object recognition files leads to low similarity matches, making it difficult to identify items accurately, and there is a lack of criteria to determine when additional registration of feature values is necessary, resulting in inefficient and labor-intensive updates during business hours.
Innovation Solution
An information processing device and method that includes a POS terminal with an imaging unit, merchandise detection, similarity calculation, and notification functions to compare item images with stored data, and notify users of the need for additional registration of collation data based on similarity rankings, allowing for real-time accuracy checks and updates.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If collation data is registered in advance for object recognition, then item identification can be performed, but recognition accuracy decreases when the registered data is improper
Solution Approach 1:
The system performs accuracy checks by comparing imaged items against stored collation data and provides feedback on recognition similarity rankings. When similarity falls below thresholds, the system notifies operators to update collation data, creating a continuous improvement loop that maintains high recognition accuracy
Solution Approach 2:
The system performs accuracy checks and similarity comparisons before final item identification is confirmed. By预先 checking the quality of collation data and predicting recognition accuracy, the system can notify operators of needed updates before they cause identification errors
2Reliability
If additional collation data registration is performed to improve recognition accuracy, then identification accuracy improves, but operational time and labor increase
Solution Approach 1:
The system automatically monitors recognition similarity and provides feedback on which collation data requires updates. This targeted feedback approach allows operators to update only specific problematic entries rather than performing comprehensive data registration, reducing time and labor
Solution Approach 2:
The system performs self-diagnosis by automatically checking recognition accuracy and identifying which collation data entries have degraded. This eliminates the need for manual, comprehensive data validation and allows the system to service itself by notifying operators only when specific updates are needed
3Measurement precision
If comprehensive accuracy checks are performed on all items, then recognition quality improves, but processing time and system complexity increase
Solution Approach 1:
The system extracts and focuses only on the critical function of checking recognition similarity between imaged items and stored collation data. By isolating this specific accuracy check function from comprehensive system validation, the patent reduces overall system complexity while maintaining essential recognition quality
Data Source
AI summary
An object recognition device includes an operation unit configured to receive a user input about an item, a storage unit that stores image data of the item, an imaging unit configured to acquire an image of the item and generate image data therefrom, and a control unit configured to compare the generated image data with the stored image data, and cause information about updating the stored image data to be presented to a user, based on a comparison result.