Handwritten Check Payee Recognition via Pixel Cluster Neural Networks
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current automated processes for processing handwritten checks fail to detect the payee, requiring users to click on the check image to view handwritten text, which is inconvenient.
Innovation Solution
A system using a neural network to identify clusters of non-white pixels in an image, applying constraints to ensure clusters are not stacked vertically and limiting their size, generating separate images for each cluster, and classifying them using machine learning to determine the characters.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If automated processes are used to process handwritten checks, then processing efficiency is improved, but the ability to detect handwritten payee information deteriorates
Solution Approach 1:
The system segments the check image into multiple regions of interest based on predefined zones (payor field, payee field, amount field, date field). Each region is processed independently by the neural network to identify and extract specific handwritten information, enabling automated detection of payee data while maintaining overall processing efficiency.
Solution Approach 2:
A neural network model serves as an intermediary between the scanned check image and the automated processing system. The neural network analyzes pixel data in specified regions, transforms visual information into structured text data, and provides this to the automated clearing house system, thereby bridging the gap between automated processing and handwritten information detection.
2Measurement precision
If cluster identification constraints are applied to prevent vertical stacking, then character recognition accuracy is improved, but processing complexity increases
Solution Approach 1:
The system applies different processing strategies to different spatial regions of the image. In regions where vertical stacking is likely (such as stacked check images), specific constraints prevent misidentification of separate characters. The neural network adjusts its clustering behavior locally based on the spatial context and predefined writing direction, improving accuracy without uniformly increasing complexity across all processing scenarios.
Data Source
AI summary
A system for identifying handwritten characters on an image using a classification model that employs a neural network. The system includes a computer having a processor and a memory device that stores data and executable code that, when executed, causes the processor to read and convert typed text on the image to machine encoded text to identify locations of the typed text on the image; identify a location on the image that includes handwritten text based on the location of predetermined typed text on the image; identify clusters of non-white pixels in the image at the location having the handwritten text, where constraints are employed to refine and limit the clusters; generate an individual and separate cluster image for each identified cluster; and classify each cluster image using machine learning and at least one neural network to determine the likelihood that the cluster is a certain character.