eKYC Document Verification Using Thickness Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing electronic Know Your Customer (eKYC) systems fail to detect illegal or falsified images of identification documents, such as driver licenses, which can be presented instead of the original, leading to potential security breaches.
Innovation Solution
A verification system that generates a transformed image of an oblique document view based on features extracted from both the front and oblique images, using algorithms like SIFT, SURF, or A-KAZE, and employs a machine learning model to determine the document's thickness and verify the authenticity of the image, ensuring the original document is used.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional text recognition is performed on uploaded images, then the eKYC process can be completed, but illegal or falsified images cannot be detected
Solution Approach 1:
The verification system divides the document image into multiple regions of interest (front surface, back surface, thickness portion) and analyzes each region separately using specialized detection algorithms. This segmentation allows the system to detect authenticity features without requiring analysis of the entire image, improving reliability while managing complexity.
Solution Approach 2:
The system transitions from two-dimensional image analysis to three-dimensional verification by capturing and analyzing the thickness dimension of the document. By measuring the thickness portion and comparing it with reference data, the system adds a dimensional check that distinguishes original documents from flat copies or falsifications.
2Measurement precision
If only front image analysis is performed, then processing is simple, but thickness and back surface authenticity cannot be verified
Solution Approach 1:
The system performs preliminary region specification to identify the thickness portion and back surface area before conducting detailed analysis. By pre-defining these regions based on geometric characteristics and reference data, the system enables precise thickness measurement without significantly increasing overall processing time.
3Ease of operation
If feature-based transformation is applied to oblique images, then document orientation is corrected, but processing time increases
Solution Approach 1:
The system extracts key geometric features (corners, edges, perspective distortion points) from the oblique image and uses only these extracted features to calculate the transformation matrix. By taking out only the essential features rather than processing the entire image, the system achieves effective orientation correction with reduced computational time.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A verification system, a verification method, and a program capable of strictly performing electronic know your customer are provided. A transformed image generation unit (32) generates a transformed image by transforming an oblique image. A target region specifying unit (36) specifies a target region indicating the target in the front image based on a feature extracted from a typical portion of a document written on a target in a sample image and a feature extracted from at least a part of the front image. A verification unit (38) determines whether a target shown in the oblique image is the same as a target shown in the front image by verifying the target region in the front image and a region in the transformed image associated with the target region.