POS Terminal Commodity Recognition via Segmented Dictionary Matching
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional POS terminal apparatuses face inefficiencies in checkout processing due to the difficulty in attaching barcodes to perishable items like vegetables and fruits, leading to longer processing times as they rely on sequential dictionary matching in object recognition, which increases with the size of the commodity dictionary.
Innovation Solution
The POS terminal apparatus employs an image recognition system that captures images of commodities, extracts feature amounts, and compares them with pre-registered dictionaries, prioritizing commodities based on shopping lists and history to streamline the recognition process.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If sequential matching processing is carried out in all recognition dictionaries, then comprehensive commodity recognition is achieved, but processing time increases with dictionary size
Solution Approach 1:
The patent divides the commodity dictionary into multiple groups (first group, second group, third group, etc.) based on different categories or priorities. The control section performs matching processing in a predetermined sequence through these groups, rather than sequentially processing all dictionaries. This segmentation reduces the time for matching processing while ensuring comprehensive commodity recognition by covering all groups in sequence.
Solution Approach 2:
The patent pre-establishes a predetermined sequence for processing different dictionary groups, and pre-organizes commodities into structured groups. This preliminary organization allows the system to efficiently navigate through the dictionary structure during matching, reducing processing time while maintaining recognition accuracy.
2Measurement precision
If comprehensive dictionary matching is performed, then all commodities are recognized, but checkout efficiency decreases
Solution Approach 1:
The patent segments the commodity dictionary into multiple groups and processes them in a predetermined sequence. This segmentation allows the system to maintain comprehensive recognition capability while reducing the time spent on matching, thereby improving checkout efficiency without sacrificing identification accuracy.
Solution Approach 2:
The patent implements dynamic processing by adjusting the matching sequence and stopping criteria based on the recognition results. The control section can terminate the matching process early when a commodity is successfully identified, or continue through additional groups if needed, optimizing the balance between accuracy and efficiency for each specific case.
Data Source
AI summary
In accordance with one embodiment, a POS terminal apparatus comprises a first interface and a control section. The first interface receives image data obtained by capturing a commodity purchased by a customer. The control section extracts a commodity belonging to a first group from a plurality of pre-registered commodities, compares feature amount of each commodity in the first group with that of an object in the image data, and compares, if a commodity corresponding to the object in the image data is not in the first group, feature amount of a commodity not belonging to the first group with that of the object.