Enhanced authentication using a secure document

Patent metadata

Field	Value
Publication number	US-12457208-B2
Application number	US-202318328957-A
Country	US
Kind code	B2
Filing date	Jun 5, 2023
Priority date	Jun 5, 2023
Publication date	Oct 28, 2025
Grant date	Oct 28, 2025

Abstract

In some implementations, a device may obtain a document that includes an identification image depicting a face of a person associated with the document. The device may obtain a live user image that depicts an image of a user. The device may extract information from the document image that includes appearance information associated with the person. The appearance information may include one or more document appearance parameters. The device may analyze the identification image, the live user image, and the appearance information to determine whether the user is the person to which the document is issued. The device may perform an action based on determining whether the user is the person to which the document is issued.

First claim

What is claimed is:

1. A system for enhanced authentication using image analysis of an identification document and appearance information associated with the identification document, the system comprising:
one or more memories; and
one or more processors, communicatively coupled to the one or more memories, configured to:
detect an authentication event associated with an access attempt for an account;
obtain a document image based on detecting the authentication event, the document image being an image of the identification document;
obtain a live user image associated with the authentication event based on detecting the authentication event, the live user image being a live image of a user;
extract information from the document image, the information including the appearance information associated with the identification document, the appearance information including one or more document appearance parameters;
analyze, using a machine learning model, the document image, the live user image, and the appearance information to determine a confidence score that indicates a likelihood that the user is associated with the identification document, the analyzing including determining whether the live user image depicts the one or more document appearance parameters, and the determining the confidence score comprises:
identifying an issue date associated with the identification document from the document image;
identifying an identification image from the document image, the identification image being an image of an owner of the identification document;
performing an image manipulation operation to modify the identification image to a modified identification image based on an amount of time between the issue date and a current date,
wherein the modified identification image is an estimated

identification image at the current date; and
comparing the live user image to the modified identification image to determine a likelihood that the live user image and the modified identification image are associated with the user; and

authenticate the access attempt based on the confidence score satisfying a threshold.

2. The system of claim 1 , wherein the one or more processors, to analyze the document image, the live user image, and the appearance information, are configured to:
identify one or more live user image appearance parameters from the live user image;
identify one or more identification image appearance parameters from the identification image; and
compare the one or more live user image appearance parameters and the one or more identification image appearance parameters to the one or more document appearance parameters,
wherein the confidence score indicates a likelihood that the one or more live user image appearance parameters and the one or more identification image appearance parameters include the one or more document appearance parameters.

3. The system of claim 1 , wherein the one or more processors, to analyze the document image, the live user image, and the appearance information, are configured to:
identify one or more live user image appearance parameters from the live user image; and
compare the one or more live user image appearance parameters to the one or more document appearance parameters to determine a likelihood that the one or more live user image appearance parameters are included in the one or more document appearance parameters.

4. The system of claim 1 , wherein the one or more processors are further configured to:
provide feedback, to an age prediction model, indicating at least one of:
the identification image,
the modified identification image,
the live user image,
the issue date,
the current date, or
an indication that the identification image and the live user image are associated with a same user,

wherein the age prediction model is associated with predicting an age of users depicted in images, and
wherein providing the feedback is based on determining that the live user image and the modified identification image are associated with the user.

5. The system of claim 1 , wherein the confidence score includes a first confidence score indicating the likelihood that the live user image and the modified identification images are associated with the user, a second confidence score indicating a likelihood that the live user image includes the one or more document appearance parameters, and a third confidence score indicating a likelihood that the live user image matches an identification image from the document image.

6. The system of claim 1 , wherein the one or more processors, to extract information from the identification document, are configured to:
obtain an image of a machine-readable code included in the document image; and
decode the machine-readable code to obtain the information.

7. The system of claim 1 , wherein the one or more processors, to extract information from the identification document, are configured to:
analyze the document image using an optical character recognition (OCR) technique to obtain the information.

8. The system of claim 1 , wherein the one or more document appearance parameters include at least one of:
an age,
an eye color,
a gender,
a skin color,
a facial characteristic,
a weight, or
a height.

9. A method of image analysis of a document and appearance information associated with the document, comprising:
obtaining, by a device, a document image, the document being an identification document, and the document including an identification image depicting a face of a person associated with the document;
obtaining, by the device, a live user image, the live user image being a live image of a user;
extracting, by the device, information from the document image, the information including appearance information associated with the person, the appearance information including one or more document appearance parameters;
analyzing, by the device, the identification image, the live user image, and the appearance information to determine whether the user is the person to which the document is issued, the analyzing including:
comparing the identification image to the live user image,
identifying whether the one or more document appearance parameters are included in the live user image,
estimating one or more live user image values of one or more live user image appearance parameters associated with the user from the live user image,
inputting, into a machine learning model, the one or more live user image values of the one or more live user image appearance parameters and one or more document values of the one or more document appearance parameters obtained via the document image, and
obtaining, from the machine learning model, a confidence score indicating a likelihood that the user is the person to which the document is issued; and

performing, by the device, an action based on determining whether the user is the person to which the document is issued.

10. The method of claim 9 , wherein the document is associated with an authenticated user of an account, and wherein performing the action comprises:
authenticating an access attempt to the account by the user based on determining that the user is the person to which the document is issued.

11. The method of claim 9 , further comprising:
obtaining an image of a machine-readable code included in the document image;
decoding the machine-readable code; and
obtaining the identification image based on decoding the machine-readable code, wherein the identification image is a color image.

12. The method of claim 9 , further comprising:
processing the document image to identify the identification image; and

Assignees

• Capital One Services Llc

Inventors

• Pribble Jason
• Patil Swapnil
• Neighbour Erik

Classifications

• G06V30/42

  based on the type of document · CPC title

• G06V40/172

  Classification, e.g. identification · CPC title

• G06V30/10

  Character recognition · CPC title

• G06V10/82

  using neural networks · CPC title

• G06V30/40

  Document-oriented image-based pattern recognition · CPC title