TECHNIQUES FOR CHECKOUT SECURITY USING VIDEO SURVEILLANCE

• US 20140063262A1
• Filed: 08/31/2012
• Published: 03/06/2014
• Est. Priority Date: 08/31/2012
• Status: Active Grant

• Current Assignee: NCR Voyix Corporation
• Original Assignee: NCR Corporation
• Inventors: Edwards, Thomas V., Bandaru, Ganesh
• Granted Patent: US 9,311,645 B2
• Time in Patent Office: Days
• US Class Current: 348/159
• CPC Class Codes: G06Q 20/20 Point-of-sale [POS] network systems
• CPC Class Codes: G06Q 20/4016 involving fraud or risk level assessment in transaction processing
• CPC Class Codes: G06Q 30/018 Certifying business or products
• CPC Class Codes: G07G 1/0054 with control of supplementary check-parameters, e.g. weight or number of articles
• CPC Class Codes: G08B 13/19613 Recognition of a predetermined image pattern or behaviour pattern indicating theft or intrusion
• CPC Class Codes: G08B 13/2462 Asset location systems combined with EAS inventory, tracking, logistic G06Q10/00; entrance control systems G07C9/00

Abstract

Techniques for checkout security using video surveillance are provided. A customer is video tracked while in a store, the pauses made in aisles and the arm movements are recorded. Expected purchased items, based on movements and pauses by the customer within the store, are then compared to actual purchased items and a determination is made whether a checkout audit is needed for the customer.

Claims

1. A processor-implemented method programmed in a non-transitory processor-readable medium and to execute on one or more processors of a machine configured to execute the method, comprising:

receiving, at the machine, in-store tracking information for a shopper;

correlating, at the machine, aspects of the in-store tracking information with known products of a store;

comparing, at the machine, purchased items with expected items when the shopper checks out of the store; and

making, at the machine, a decision as to whether to audit the shopper before the shopper exits the store.

2. The method of claim 1, wherein receiving further includes acquiring the in-store tracking information as data showing movements of the shopper through the store, pauses of the shopper in known areas of the store, and item identifiers in the known areas of the store.

3. The method of claim 2, wherein acquiring further includes identifying in the pauses data points for arm movements of the shopper to restrict the item identifiers.

4. The method of claim 1, wherein receiving further includes obtaining the in-store tracking information initially as video data of the shopper traversing the store and captured by multiple cameras dispersed throughout the store.

5. The method of claim 4, wherein obtaining further includes passing the video data to a decision agent that returns data representing movements of the shopper through the store, pauses of the shopper in known areas of the store, arm movements of the shopper during the pauses, and item identifiers in the known areas of the store having the arm movements.

6. The method of claim 1, wherein correlating further includes accessing a store item database to resolve the expected items based on the in-store tracking information.

7. The method of claim 1, wherein comparing further includes receiving the purchased items in real time from a kiosk or cashier manned device when the shopper checks out.

8. The method of claim 1, wherein comparing further includes identifying at least one purchased item as being of a same type as an expected item indicating that the shopper misidentified the at least one purchased item.

9. The method of claim 1, wherein making further includes evaluating a variety of additional heuristics before making the decision.

10. The method of claim 1, wherein making further includes deciding to request an audit when at least one expected item is missing from the purchased items.

11. The method of claim 10, wherein deciding further includes sending an alert to in-store personnel to immediately visit the shopper and perform the requested audit.

12. A processor-implemented method programmed in a non-transitory processor-readable medium and to execute on one or more processors of a machine configured to execute the method, comprising:

receiving, at the machine, a list of expected items that a shopper within a store is expected to have at a checkout;

identifying, at the machine, purchased items that the shopper claims at the checkout; and

deciding, at the machine, whether to perform an audit of the shopper when at least one of the expected items does not exist in the purchased items at the conclusion of the checkout.

13. The method of claim 12, wherein receiving further includes acquiring the expected items from an in-store person tracking system that tracked the shopper throughout the store.

14. The method of claim 12, wherein identifying further includes acquiring the purchased items from the shopper at a self-service checkout station.

15. The method of claim 12, wherein identifying further includes acquiring the purchased items from a cashier at a cashier manned checkout station.

16. The method of claim 12, wherein deciding further includes sending an alert to someone other than a cashier involved in the checkout when the audit was decided to be performed.

17. The method of claim 16 further comprising, recording the decision to audit as being associated with the cashier.

18. A system comprising:

a machine configured with an audit manager that executes on one or more processors of the machine, and the audit manager is implemented and resides in a non-transitory computer-readable storage medium;

wherein the audit manager resolves expected items that a shopper within a store is expected to have and compares those expected items against actual purchased items at a checkout for the shopper to determine whether a transaction for the shopper is to be audited.

19. The system of claim 18, wherein the audit manager is configured to interact with an in-store people tracking system to resolve the expected items.

20. The system of claim 19, wherein the audit manager is configured to interact with an in-store product database to resolve the expected items.