Computer Vision Cart Verification for Frictionless Store Exits
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for identifying unpaid items at store exits are inefficient and create friction, leading to increased wait times and customer dissatisfaction due to random scanning of a limited number of items, resulting in potential loss and reduced customer loyalty.
Innovation Solution
A system utilizing computer vision and object detection models to identify items in customer carts in real-time by selecting optimal images based on anchor points, comparing scanned items with e-receipts, and generating notifications for unpaid items, allowing customers to quickly address any discrepancies before exiting.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If random item scanning is performed at store exit, then unpaid item detection is achieved, but customer wait time increases and exit friction increases
Solution Approach 1:
The system performs preliminary actions by capturing images of customer carts at multiple anchor points throughout the store journey. Item detection and recognition models process these images in real-time to build a complete inventory of items in each cart before the customer reaches the exit, eliminating the need for last-minute scanning and reducing wait time.
Solution Approach 2:
The patent replaces the mechanical manual scanning process with computer vision technology. Image capture devices continuously photograph carts, and AI models automatically detect and recognize items in the images, substituting the manual scanner operation with an automated visual inspection system that operates without customer interaction.
2Productivity
If only a few random items are scanned at exit, then some unpaid items are detected, but verification completeness deteriorates
Solution Approach 1:
The system maintains continuous useful action by capturing images at multiple anchor points throughout the customer's shopping journey rather than performing a single random check at exit. This continuous monitoring ensures that all items in the cart are detected and verified, achieving complete verification while maintaining high productivity through automated real-time processing.
3Reliability
If manual item scanning is performed at exit, then unpaid items can be identified, but customer experience deteriorates due to friction
Solution Approach 1:
The system implements self-service by having the computer vision technology automatically perform item detection and verification without requiring customer participation. The AI models independently analyze cart images, identify items, compare against purchase data, and flag potential unpaid items, allowing customers to proceed through exit without manual scanning interruptions.
4Reliability
If comprehensive cart verification is performed, then all unpaid items are detected, but processing complexity increases
Solution Approach 1:
The patent applies segmentation by dividing the verification process into distinct functional modules: image capture at anchor points, item detection models, item recognition models, cart identification, and unpaid item flagging. This modular segmentation allows each component to specialize in its specific task, improving overall detection accuracy while managing system complexity through organized functional separation.
Data Source
AI summary
Examples provide a system and method for identifying unpaid items in real-time using computer vision object detection and recognition models. Images of a cart are selected based on proximity of the cart to one or more anchor points. The object detection and recognition models analyze the image data and identify a set of items in the selected cart. An e-receipt including a set of paid items is selected from a plurality of active electronic receipts based on matching the set of identified items to the set of paid items. Any unmatched items are added to a set of predicted unpaid items. When a receipt corresponding to the selected e-receipt is scanned, a notification identifying the set of predicted unpaid items is provided to a user device for display, enabling a user to identify any unpaid items in a basket of items quickly and accurately in real-time at store exit.