Large Basket Vision Checkout With Duplicate Item Tracking
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing vision-based self-service checkout systems are limited to processing a single basket or tray of items at a time, leading to inefficiencies and the need for manual intervention when handling larger quantities.
Innovation Solution
A system and method that allows multiple trays of items to be processed by comparing current and previous item lists, enabling automatic detection and correction of duplicates, and using a heartbeat mechanism to maintain system operation, without requiring manual intervention.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If a vision system processes only a single basket or tray at a time, then the system complexity remains manageable, but the productivity and efficiency of the checkout process deteriorates when handling larger quantities of items
Solution Approach 1:
The system segments the checkout process by introducing multiple virtual trays that can be independently tracked and processed. Each physical basket is divided into multiple virtual trays, allowing the vision system to process items in smaller chunks while maintaining overall productivity. This segmentation enables the system to handle large quantities of items without requiring a single complex processing unit.
Solution Approach 2:
The patent introduces a temporal dimension to the processing by implementing a queue system where multiple virtual trays are processed sequentially. This transforms the single-dimension (single tray) processing into a multi-dimensional approach where trays are organized in a queue, allowing the system to maintain simple per-tray processing logic while achieving high overall throughput through parallel queue management.
2Ease of operation
If manual intervention is required for large baskets, then accuracy can be maintained, but the ease of operation and convenience deteriorates
Solution Approach 1:
The system implements self-service by automatically detecting when items are placed on or removed from trays using vision technology. The duplicate detection and correction mechanisms operate autonomously without requiring manual verification, allowing customers to simply place items on the conveyor while the system handles tracking, validation, and correction processes automatically.
Solution Approach 2:
The system employs continuous feedback loops where the vision system monitors item placement, the processor detects potential duplicates by comparing against the item list, and the system provides real-time feedback to the customer display. This feedback mechanism maintains accuracy by allowing customers to verify and correct item tracking automatically during the checkout process.
3Reliability
If duplicates are not automatically detected and corrected, then the device complexity remains low, but the reliability of the checkout process deteriorates
Solution Approach 1:
The system performs preliminary duplicate detection by maintaining a running item list and comparing each newly detected item against this list before finalizing the transaction. This preliminary action prevents duplicates from being processed, ensuring transaction accuracy early in the checkout process rather than requiring complex post-processing verification.
Solution Approach 2:
The patent replaces manual verification mechanisms with automated vision-based detection and processor-based duplicate identification. Instead of requiring customers or operators to manually check for duplicates, the system uses computer vision to track items and algorithmic processing to identify duplicates, substituting mechanical/manual processes with automated electronic systems that improve reliability.
Data Source
AI summary
A self-service checkout system has a processor and a non-transitory computer-readable storage medium storing executable instructions. The executable instructions, when executed by the processor, cause the processor to perform operations including: receiving a message from a vision system that includes a current list of identified items on a tray positioned adjacent to the vision system for a current transaction, comparing the current list of identified items to a previous list of identified items, and adding any items on the current list that are not on the previous list to a transaction list for a current transaction, determining that the current transaction is complete when the current list of identified items has no items, enabling a customer to review the transaction list and remove any duplicate entries, and completing the current transaction.