Document Image Quality Assessment for Faster Authentication
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Traditional document image analysis systems often reject images that are of sufficient quality due to imperfections, leading to inefficiencies and delays in authentication processes, and existing techniques require multiple stages and different algorithms for various quality issues.
Innovation Solution
A convolutional neural network (CNN) with inception and self-attention blocks is trained on a small labeled dataset to automatically evaluate image quality by focusing on important document areas, identifying and tolerating imperfections that do not obscure critical information.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional document image analysis systems use multiple stages and different algorithms to assess image quality, then measurement precision is improved, but device complexity increases
Solution Approach 1:
The patent combines multiple quality assessment algorithms and stages into a single unified deep learning model that simultaneously evaluates multiple quality dimensions (blur, glare, reflection, obstruction, resolution) in one pass, reducing system complexity while maintaining comprehensive assessment capability
Solution Approach 2:
The deep learning model serves multiple functions: it detects various types of imperfections (blur, glare, reflection, obstruction), assesses overall image quality, and determines authentication suitability, replacing the need for separate specialized algorithms for each quality aspect
2Reliability
If traditional systems reject images with imperfections, then reliability is improved, but productivity decreases
Solution Approach 1:
The system changes the decision parameter from binary reject/accept based on simple quality thresholds to a nuanced assessment that weighs the severity and location of imperfections against their impact on critical information, allowing acceptable images to pass while maintaining security
Solution Approach 2:
The model applies different quality standards to different regions of the image, tolerating imperfections in non-critical areas while maintaining strict requirements for regions containing critical authentication information, thereby reducing unnecessary rejections
3Measurement precision
If manual review is used to assess document image quality, then measurement precision is improved, but loss of time increases
Solution Approach 1:
The patent replaces the mechanical manual review process with an automated deep learning-based image quality assessment system that performs comprehensive quality evaluation in seconds, eliminating the time-consuming human review while maintaining or improving assessment consistency
4Reliability
If operators perform manual review to verify document information, then reliability is improved, but ease of operation worsens
Solution Approach 1:
The system performs self-verification by automatically assessing image quality and determining authentication suitability without requiring operator intervention, reducing workload while maintaining verification standards through automated decision-making
Data Source
AI summary
Techniques are disclosed relating to automatically determining image quality for images of documents. In some embodiments, a computer system receives an image of a document captured at a user computing device. Using a neural network, the computer system analyzes the image to determine whether the image satisfies a quality threshold, where the analyzing includes determining whether one or more features in the image used in an authentication process are obscured. The computer system transmits, to the user computing device, a quality result, where the quality result is generated based on an image classification output by the neural network. Automatically determining whether a received image of a document satisfies a quality threshold may advantageously improve the chances of a system being able to complete an authentication process quickly, which in turn may improve user experience while reducing fraudulent activity.