Flat Card Scanning Using Machine Learning for Non-Embossed Characters
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing card scanning technologies struggle to accurately read information from cards with non-embossed characters and varied designs, limiting their usability in mobile applications.
Innovation Solution
A system utilizing a machine learning model, including a neural network, to detect account numbers and user identifiers on cards regardless of font, location, or format, enabling scanning of any type of card and storing the information in a virtual wallet.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional card scanning technologies are used, then scanning of cards with standard designs can be performed, but scanning accuracy deteriorates for cards with non-embossed characters and varied designs
Solution Approach 1:
The system transforms the card image through multiple parameter changes including grayscale conversion, thresholding, and noise filtering to enhance character visibility. These parameter transformations enable the OCR engine to accurately detect non-embossed characters and adapt to various card designs, fonts, and layouts, thereby resolving the contradiction between scanning accuracy and design versatility
Solution Approach 2:
The patent replaces traditional mechanical/optical scanning methods with a machine learning-based image processing system. The neural network model automatically learns and adapts to different card designs, fonts, and character styles, substituting the need for multiple specialized scanning mechanisms with a single adaptive software-based system that maintains high accuracy across diverse card types
2Adaptability or versatility
If a machine learning model is applied to detect card information, then adaptability to various card designs is improved, but processing time increases
Solution Approach 1:
The system performs preliminary image preprocessing operations including grayscale conversion, thresholding, and noise filtering before applying the machine learning model. This preliminary action prepares the image data in an optimized format that reduces the computational burden on the neural network, thereby decreasing processing time while maintaining the model's adaptability to various card designs
Solution Approach 2:
The card scanning process is segmented into distinct stages: image capture, preprocessing, character detection, and information extraction. Each segment is optimized independently, with the preprocessing stage handling routine transformations and the machine learning model focusing only on the complex task of recognizing diverse character styles, thereby reducing overall processing time while preserving adaptability
Data Source
AI summary
The present disclosure is directed to systems and methods that enable scanning of any type of card regardless of the shape and design of a given card and/or a font, a shape and a format with which characters such as numbers, letters and symbols are printed on the cards including cards with non-embossed characters printed thereon. In one example, a method includes scanning a card, the card including at least an account number associated with a user of the card and an identifier of the user; detecting, by applying a machine learning model to the card after scanning the card, at least the account number printed on the card; and completing a task using the account number.