Client Side Card OCR Image Filtering
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current systems for extracting financial card information from mobile devices are cumbersome due to small screen sizes and keyboard interfaces, leading to errors and the need for precise image capture for optical character recognition (OCR), with users often transmitting multiple images before obtaining a usable one, which burdens processing capacity.
Innovation Solution
Implementing a user computing device that filters images before transmission to an OCR system by analyzing features such as blurriness, brightness, and text presence, adjusting image capture methods, and repeating the process until a usable image meets configured thresholds for OCR processing.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If users transmit multiple images to the OCR system before obtaining a usable one, then the likelihood of obtaining a usable image increases, but the processing capacity burden increases
Solution Approach 1:
The system performs preliminary filtering of images on the user device before transmission to the OCR system. Image quality metrics such as blur detection, text visibility assessment, and lighting evaluation are conducted locally to pre-screen images, ensuring only potentially usable images are transmitted for OCR processing.
Solution Approach 2:
The user device autonomously evaluates image quality using onboard processing capabilities. The device independently determines whether captured images meet minimum quality thresholds without requiring server-side pre-processing, enabling self-service filtering that reduces unnecessary transmissions to the OCR system.
2Measurement precision
If users are required to precisely position the card in the scan box, then the accuracy of image capture improves, but the ease of operation deteriorates
Solution Approach 1:
The scanning box is implemented as a dynamic, flexible guide rather than a rigid fixed boundary. The system adapts the scanning region based on detected card features and image quality, allowing the box to resize and reposition automatically. This dynamic adjustment maintains measurement precision while reducing the need for manual user intervention to achieve precise positioning.
Solution Approach 2:
The system provides real-time feedback to users regarding image quality and positioning accuracy. Visual indicators show whether the card is properly positioned and whether image quality metrics (blur, lighting, text visibility) meet thresholds. This feedback loop guides users to make minimal adjustments rather than requiring perfect initial positioning.
3Reliability
If the system processes multiple images without filtering, then the chance of finding a usable image increases, but the loss of time increases
Solution Approach 1:
Image quality assessment is performed in advance on the user device before images are transmitted to the OCR system. Metrics such as blur detection, text visibility, and lighting evaluation are computed locally to filter out unusable images beforehand, preventing wasteful transmission and processing of low-quality images.
Solution Approach 2:
The system replaces manual user review and selection of images with automated computer-based image quality assessment algorithms. Machine learning models and computer vision techniques evaluate image suitability objectively and rapidly, substituting time-consuming manual inspection with efficient automated filtering that maintains high reliability while reducing time loss.
Data Source
AI summary
The technology of the present disclosure includes computer-implemented methods, computer program products, and systems to filter images before transmitting to a system for optical character recognition (“OCR”). A user computing device obtains a first image of the card from the digital scan of a physical card and analyzes features of the first image, the analysis being sufficient to determine if the first image is likely to be usable by an OCR algorithm. If the user computing device determines that the first image is likely to be usable, then the first image is transmitted to an OCR system associated with the OCR algorithm. Upon a determination that the first image is unlikely to be usable, a second image of the card from the digital scan of the physical card is analyzed. The optical character recognition system performs an optical character recognition algorithm on the filtered card.