Biometric Database Segmentation for Fraud Detection

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Large databases of biometric images, such as those used in motor vehicle registries, face challenges with high false positives and processing delays due to their size, making it difficult to effectively detect identity theft and fraud, especially when databases contain millions of records and lack operator intervention during image capture.

Innovation Solution

A browser-based system with an operator-friendly interface for searching databases of captured images using facial recognition techniques like local feature analysis (LFA), which allows for automated biometric searching, image alignment, and fraud detection by comparing probe images against large databases of biometric templates.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If automated biometric searching is implemented in large databases, then fraud detection capability is improved, but processing time increases

Engineering Contradiction:

Improvefraud detection capabilityVSAvoidprocessing time

Core Design Contradiction:

ReliabilityVSLoss of time

Solution Approach 1:

The system segments the biometric database into multiple smaller databases organized by geographic region or jurisdiction. Instead of searching the entire large database at once, the probe image is searched against relevant segments only, significantly reducing processing time while maintaining comprehensive fraud detection coverage across the entire database.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The system performs preliminary actions by pre-processing and organizing database records into structured segments with metadata indexing before actual search operations. This pre-organization allows for rapid retrieval and comparison of relevant records during fraud detection, reducing real-time processing requirements.

Inventive Principle:

Principle #10Preliminary action

2Measurement precision

If comprehensive database searching is performed, then fraud detection accuracy is improved, but false positives increase

Engineering Contradiction:

Improvefraud detection accuracyVSAvoidfalse positives

Core Design Contradiction:

Measurement precisionVSObject-generated harmful factors

Solution Approach 1:

The system applies local quality by searching and comparing biometric data within specific local segments or regions rather than uniformly across the entire database. This targeted approach allows for more precise matching by considering local characteristics and reducing the impact of irrelevant data, thereby decreasing false positives while maintaining detection accuracy.

Inventive Principle:

Principle #3Local quality

Solution Approach 2:

The system incorporates feedback mechanisms where search results and detection outcomes are continuously reviewed and used to refine future search parameters. This feedback loop allows the system to learn from previous detections and adjust its criteria, reducing false positives over time while improving overall detection accuracy.

Inventive Principle:

Principle #23Feedback

3Ease of operation

If manual image capture and review is used, then operator control is maintained, but processing efficiency decreases

Engineering Contradiction:

Improveoperator controlVSAvoidprocessing efficiency

Core Design Contradiction:

Ease of operationVSProductivity

Solution Approach 1:

The system implements self-service by enabling automated biometric comparison and fraud detection without requiring constant manual intervention. The automated system performs image capture, biometric extraction, database searching, and initial fraud detection independently, maintaining operator control through supervisory access while dramatically improving processing efficiency.

Inventive Principle:

Principle #25Self-service

Solution Approach 2:

The system introduces an intermediary automated biometric processing layer between image capture and final fraud determination. This intermediary system handles the bulk of processing tasks automatically, with operators intervening only for review and complex cases, thus maintaining operational control while enhancing processing efficiency.

Inventive Principle:

Principle #24Intermediary (Mediator)

Data Source

PatentUS7804982B2Systems and methods for managing and detecting fraud in image databases used with identification documents

Publication Date: 2010.09.28 IDEMIA CIVIL IDENTITY NA LLC

AI summary

We provide a system for issuing identification documents to a plurality of individuals, comprising a first database, a first server, and a workstation. The first database stores a plurality of digitized images, each digitized image comprising a biometric image of an individual seeking an identification document. The first server is in operable communication with the first database and is programmed to send, at a predetermined time, one or more digitized images from the first database to a biometric recognition system, the biometric recognition system in operable communication with a second database, the second database containing biometric templates associated with individuals whose images have been previously captured, and to receive from the biometric recognition system, for each digitized image sent, an indicator, based on the biometric searching of the second database, as to whether the second database contains any images of individuals who may at least partially resemble the digitized image that was sent. The a workstation is in operable communication with the first server and is configured to permit a user to review the indicator and to make a determination as to whether the individual is authorized to be issued an identification document or to keep an identification document in the individual's possession.