Counterfeit detection using image analysis

Patent metadata

Field	Value
Publication number	US-12437573-B2
Application number	US-202318301372-A
Country	US
Kind code	B2
Filing date	Apr 17, 2023
Priority date	Apr 17, 2023
Publication date	Oct 7, 2025
Grant date	Oct 7, 2025

Abstract

In some implementations, a device may obtain a first image of a first note, and identify a first identifier associated with the first note and a first set of visual characteristics of the first note indicating an appearance of the first note. The device may obtain a second image of a second note, and identify a second identifier associated with the second note and a second set of visual characteristics of the second note indicating an appearance of the second note. The second identifier may correspond to the first identifier, indicating that the second note is purported to be the first note. The device may determine whether the second note is counterfeit based on the first set of visual characteristics of the first note and the second set of visual characteristics of the second note. The device may perform action(s) based on a determination that the second note is counterfeit.

First claim

What is claimed is:

1. A system for counterfeit detection using image analysis, the system comprising:
one or more memories; and
one or more processors, communicatively coupled to the one or more memories, configured to:
obtain a first image of a first note that is provided to a user from an entity;
identify, based on the first image, a first identifier associated with the first note and a first set of visual characteristics of the first note indicating an appearance of the first note;
obtain a second image of a second note that is incoming to the entity;
identify, based on the second image, a second identifier associated with the second note and a second set of visual characteristics of the second note indicating an appearance of the second note,
wherein the second identifier corresponds to the first identifier indicating that the second note is purported to be the first note;

determine, using a machine learning model, whether the second note is counterfeit based on the first set of visual characteristics of the first note, the second set of visual characteristics of the second note, and at least one of an occupation of the user or an interaction history associated with the user; and
transmit, based on a determination that the second note is counterfeit, an indication that the second note is counterfeit.

2. The system of claim 1 , wherein the one or more processors are further configured to:
generate information indicating an association between the first identifier, the first set of visual characteristics of the first note, and the user.

3. The system of claim 1 , wherein the one or more processors are further configured to:
retrieve, based on the first identifier corresponding to the second identifier, information indicating the first set of visual characteristics of the first note and data indicating the at least one of the occupation of the user or the interaction history associated with the user.

4. The system of claim 1 , wherein the one or more processors, to determine, using the machine learning model, whether the second note is counterfeit, are configured to:
determine, using the machine learning model, whether the second note is counterfeit based on the first set of visual characteristics of the first note, the second set of visual characteristics of the second note, the at least one of the occupation of the user or the interaction history associated with the user, a first time of the first note being provided to the user, a second time of the second note incoming to the entity, a first location of the first note being provided to the user, and a second location of the second note incoming to the entity.

5. The system of claim 1 , wherein the first set of visual characteristics of the first note relate to damage to the first note, a coloration of the first note, or a quality of the first note, and
wherein the second set of visual characteristics of the second note relate to damage to the second note, a coloration of the second note, or a quality of the second note.

6. The system of claim 1 , wherein the one or more processors are further configured to:
perform optical character recognition on at least a portion of the first image that is associated with the first identifier; and
perform optical character recognition on at least a portion of the second image that is associated with the second identifier.

7. The system of claim 1 , wherein the one or more processors are further configured to:
cause a note handling device to reject a deposit of the second note based on the determination that the second note is counterfeit.

8. The system of claim 1 , wherein the interaction history indicates one or more historical interactions between the user and one or more merchants in connection with an account of the user maintained by the entity.

9. A method of counterfeit detection using image analysis, comprising:
obtaining, by a device, a first image of a first note provided to a user;
identifying, by the device and based on the first image, a first identifier associated with the first note and a first set of visual characteristics of the first note indicating an appearance of the first note;
obtaining, by the device and subsequent to the first note being provided to the user, a second image of a second note;
identifying, by the device and based on the second image, a second identifier associated with the second note and a second set of visual characteristics of the second note indicating an appearance of the second note,
wherein the second identifier corresponds to the first identifier indicating that the second note is purported to be the first note;

determining, by the device and using a machine learning model, whether the second note is counterfeit based on the first set of visual characteristics of the first note, the second set of visual characteristics of the second note, and data associated with the user; and
transmitting, by the device and based on a determination that the second note is counterfeit, an indication that the second note is counterfeit.

10. The method of claim 9 , wherein the data associated with the user indicates an occupation of the user.

11. The method of claim 9 , wherein the data associated with the user indicates an interaction history of the user, and
wherein the interaction history indicates one or more historical interactions between the user and one or more merchants in connection with an account of the user.

12. The method of claim 9 , wherein the second note is from a different user, and
wherein determining, using the machine learning model, whether the second note is counterfeit comprises:
determining, using the machine learning model, whether the second note is counterfeit based on the first set of visual characteristics of the first note, the second set of visual characteristics of the second note, the data associated with the user, and additional data associated with the different user.

13. The method of claim 9 , wherein the first set of visual characteristics of the first note relate to damage to the first note, a coloration of the first note, or a quality of the first note, and
wherein the second set of visual characteristics of the second note relate to damage to the second note, a coloration of the second note, or a quality of the second note.

14. The method of claim 9 , wherein the machine learning model is trained to classify a note into a first classification indicating that the note is counterfeit or into a second classification indicating that the note is authentic.

15. A non-transitory computer-readable medium storing a set of instructions for counterfeit detection using image analysis, the set of instructions comprising:
one or more instructions that, when executed by one or more processors of a device, cause the device to:
obtain a first image of a first note provided to a user;
identify, based on the first image, a first identifier associated with the first note and a first set of visual characteristics of the first note indicating an appearance of the first note;
obtain a second image of a second note;
identify, based on the second image, a second identifier associated with the second note and a second set of visual characteristics of the second note indicating an appearance of the second note,
wherein the second identifier corresponds to the first identifier indicating that the second note is purported to be the first note;

determine whether the second note is counterfeit based on the first set of visual characteristics of the first note and the second set of visual characteristics of the second note; and
perform

Assignees

• Capital One Services Llc

Inventors

• Yip Samuel
• Garcia Mario A
• Irizarry Roberto

Classifications

• G06V30/418 Primary

  Document matching, e.g. of document images · CPC title

• G06V10/82

  using neural networks · CPC title

• G06V30/42 Primary

  based on the type of document · CPC title