Document Verification Using Neural Network Segmentation

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Current remote identity document verification systems face challenges in accurately distinguishing physical documents from backgrounds and verifying their authenticity, especially in uncontrolled environments, leading to potential fraudulent activities.

Innovation Solution

The method employs an artificial neural network for semantic segmentation to separate physical documents from backgrounds, combined with corner detection and brightness analysis, and uses a machine learning classifier to verify document authenticity based on predefined features, leveraging a fully convolutional neural network trained on an augmented dataset with geometric transformations.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Ease of operation

If remote ID verification systems use simple image extraction methods, then the system is easy to operate, but the accuracy of distinguishing physical documents from backgrounds deteriorates in uncontrolled environments

Engineering Contradiction:

Improveease of operationVSAvoiddocument separation accuracy

Core Design Contradiction:

Ease of operationVSMeasurement precision

Solution Approach 1:

The verification system segments the document image from the background using corner detection algorithms that identify characteristic corner points of the document. This segmentation separates the document regions from uncontrolled backgrounds without requiring complex manual processing, maintaining ease of operation while improving separation accuracy.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The system replaces manual document verification with automated image processing algorithms including corner detection, brightness analysis, and geometric transformation. This substitution of mechanical/manual processes with computational methods enables accurate document-background separation in uncontrolled environments while keeping the system easy to operate.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

2Device complexity

If the system uses basic verification methods, then the device complexity is low, but the reliability of document authenticity verification deteriorates

Engineering Contradiction:

Improveverification system complexityVSAvoidauthenticity verification reliability

Core Design Contradiction:

Device complexityVSReliability

Solution Approach 1:

The system performs preliminary actions by detecting corner points and analyzing brightness distributions before conducting authenticity verification. These preliminary steps prepare the image data and identify key features, enabling reliable verification without requiring excessively complex processing during the main verification phase.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The verification system uses feedback mechanisms where extracted document features are compared against expected characteristics, and the results feed into subsequent verification steps. This feedback loop improves reliability by continuously validating document authenticity based on multiple extracted features including corner positions and brightness patterns.

Inventive Principle:

Principle #23Feedback

3Measurement precision

If the system performs comprehensive document analysis, then the authenticity verification accuracy is improved, but the processing time increases leading to more inconclusive results

Engineering Contradiction:

Improveverification accuracyVSAvoidprocessing time

Core Design Contradiction:

Measurement precisionVSLoss of time

Solution Approach 1:

The system applies partial action by focusing analysis on key document regions identified through corner detection rather than processing the entire image uniformly. Brightness analysis and feature extraction are concentrated on identified document areas, achieving high verification accuracy while reducing overall processing time compared to comprehensive full-image analysis.

Inventive Principle:

Principle #16Partial or excessive action

Data Source

PatentUS11144752B1Physical document verification in uncontrolled environments

Publication Date: 2021.10.12 EASY SOLUTIONS ENTERPRISES CORP

AI summary

A method for verifying authenticity of a physical document includes receiving an image of a physical document to be authenticated including the physical document and a background. A pre-processed image is produced that includes the physical document separated from the background. The producing includes separating the physical document from the background by semantic segmentation utilizing an artificial neural network trained using an augmented dataset generated by applying geometric transformations over different backgrounds. Features of the pre-processed image are extracted to determine a document type. In response to determining the document type of the physical document, the method includes verifying, utilizing a machine learning classifier, whether the physical document is authentic based on the extracted features relative to expected features for the corresponding document type. An indication of whether the physical document is authentic based on the verifying is generated.