Authentication of a suspect object using extracted native features

Patent metadata

Field	Value
Publication number	US-10915749-B2
Application number	US-201816041710-A
Country	US
Kind code	B2
Filing date	Jul 20, 2018
Priority date	Mar 2, 2011
Publication date	Feb 9, 2021
Grant date	Feb 9, 2021

Abstract

A forgery detection system includes a computer server and a database system of digital fingerprint records corresponding to forged or altered objects of a given object type. Using the computer server, a digital image of a suspect object of the given object type is accessed, an authentication region is selected, and a native feature within the authentication region is extracted. The native feature describes physical characteristics of the authentication region without recognizing content that appears in the authentication region. The computer server forms a feature vector to represent the native feature in a compact form and queries the database system to obtain a result responsive to digital fingerprint records that match the feature vector. Each matching digital fingerprint record is counted, and if the count of fraud indicator matches crosses a predetermined threshold indicating a confidence level that the suspect object is forged or altered, a report based is generated and communicated to a user interface.

First claim

The invention claimed is:

1. A forgery detection system, comprising:
a computer server;
a database system operatively coupled to the computer server, the database system storing a forgery database comprising multiple digital fingerprint records, each of the digital fingerprint records having at least one feature vector extracted from a forged or altered object of a given object type; and
instructions stored in non-volatile memory and executable in the computer server to cause the computer server to:
access a digital image of a suspect object of the given object type, wherein the digital image of the suspect object is captured by a scanner, a camera, a specially-adapted sensor array such as CCD array, a microscope, a smart phone camera, or video camera;
select an authentication region of the digital image;
extract a native feature within the selected authentication region wherein the native feature describes physical characteristics of the selected authentication region without recognizing content that appears in the selected authentication region;
form a feature vector to represent the native feature in a compact form;
query the database system based on the feature vector to obtain a result responsive to zero or more of the digital fingerprint records stored in the database system matching the feature vector, each matching digital fingerprint record defining a fraud indicator match;
for each defined fraud indicator match, incrementing a count of fraud indicator matches;
compare the count of fraud indicator matches to a predetermined threshold value to determine a confidence level that the suspect object is forged or altered;
generate a report based on the comparison; and
transmit the report to a user interface.

2. The system of claim 1 wherein:
at least some of the multiple digital fingerprint records of the forgery database include plural features from a single digital image; and
for at least some of the plural features, the stored instructions cause the computer server to:
extract an additional feature from the single digital image;
query the database system to find digital fingerprint records that match the additional feature;
increment the count of fraud indicator matches; and
update the determined confidence level that the suspect object is forged or altered.

3. The system of claim 1 wherein:
the forgery database comprises digital fingerprint records that include at least one corresponding feature from each one of multiple authentication regions; and
the stored instructions cause the computer server to:
select a first authentication region based on the given object type;
extract a first feature from the first authentication region;
query the database system to identify first digital fingerprint records that match the first authentication region and first feature data pair;
select a second authentication region based on the given object type;
extract a second feature from the second authentication region;
query the database system to identify second digital fingerprint records that match the second authentication region and second feature data pair; and
include a count of both first and second matching records in the count of fraud indicator matches.

4. The system of claim 1 wherein the given object type is one of a negotiable instrument as defined in the Uniform Commercial Code, a bill of lading or shipping document, an original work of art, a label on a bottle of wine.

5. The system of claim 1 wherein selecting the authentication region is based on a predetermined template accessible to the computer server that defines at least one authentication region for the given object type.

6. The system of claim 1 wherein at least some of the multiple digital fingerprint records stored in the database system include data that indicates a source of the corresponding forged or altered object; and
the stored instructions executable in the computer server further to cause the computer server to:
generate or include in the report to the user interface an indication of the source(s) of the matching record(s) to help identify persons or entities that may have forged or altered the suspect object.

7. The system of claim 1 wherein the stored instructions executable in the computer server further to cause the computer server to select the predetermined threshold value for determining the confidence level based on the given object type.

8. The system of claim 1 wherein the confidence level is determined empirically.

9. The system of claim 1 wherein the stored instructions executable in the computer server further to cause the computer server, in a case that the result returns no matching record, to update the digital fingerprint of the suspect object to reflect that no match was found.

10. The system of claim 1 wherein the stored instructions executable in the computer server further to cause the computer server, in a case that the result returns a matching record, to update the digital fingerprint of the matching record to reflect that the match was found.

11. The system of claim 10 wherein the stored instructions executable in the computer server further to cause the computer server to update the digital fingerprint of the matching record to include a link to the suspect object or a record associated with the suspect object.

12. A method, comprising:
acquiring digital image data from at least a selected authentication region of an object, without printing or adding anything on to the object, and wherein the object has been previously forged or altered at least in the selected authentication region from its original state;
extracting at least one native feature from the acquired digital image data;
wherein the at least one native feature describes physical characteristics of the selected authentication region without recognizing content that appears in the selected authentication region;
forming a feature vector to represent the native feature in a compact form;
storing the feature vector in a database record as part of a digital fingerprint of the object in a forgery database for use in detecting forged or altered objects among a class of objects to which the forged or altered object belongs;
repeating the foregoing acquiring, extracting, forming and storing over additional forged or altered objects of the class of objects;
based on the repeating, adding additional records to the forgery database, the additional records comprising feature vectors representing corresponding native features of the selected authentication region in the additional forged or altered objects;
acquiring second digital image data of a portion of a suspect object, wherein the portion of the suspect object corresponds to the selected authentication region used to build the forgery database;
extracting second features from the second image data;
forming a second feature vector to represent the second features in the compact form;
querying the forgery database using the second feature vector to obtain a result based on zero or more stored feature vectors that match the second feature vector, each stored feature vector that matches the second feature vector defining a fraud indicator match;
counting a number of fraud indicator matches;
comparing the number of fraud indicator matches to a predetermined threshold value to determine a confidence level that the suspect object is forged or altered;
generating a report of the determined confidence level based on the comparison; and
transmitting the report to a user interface.

13. The method of claim 12 wherein each feature vector comprises an array of color or gray scale numeric va

Assignees

• Alitheon Inc

Inventors

• Ross David
• Elmenhurst Brian
• Tocci Mark
• Forbes John
• Ross Heather Wheelock

Classifications

• G06V30/424 Primary

  Postal images, e.g. labels or addresses on parcels or postal envelopes · CPC title

• G06V30/418

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

• G06V10/245

  by locating a pattern; Special marks for positioning · CPC title

• G06V30/10

  Character recognition · CPC title

• G06T7/90

  Determination of colour characteristics · CPC title