Voucher Verification Using Image Feature Vectors and OCR
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The inefficiency in verifying agreement between accounting data and paper-based vouchers due to varying voucher formats and noise in optical character recognition (OCR) outputs, leading to incomplete and inaccurate character strings, necessitates a method for automated verification regardless of OCR performance.
Innovation Solution
A voucher verification method and system that utilize a processing portion to extract local image features and text data from vouchers, generate vectors using a learned determination model, and calculate similarity to determine agreement between accounting data and vouchers, enabling automated verification regardless of OCR accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Extent of automation
If OCR is used to extract character strings from voucher images, then automated data extraction is achieved, but the extracted character strings contain noise and are incomplete, reducing accuracy
Solution Approach 1:
The patent segments the voucher image into multiple local regions and extracts image feature values from each region. This segmentation allows the system to capture spatial relationships and contextual information that OCR alone cannot provide, thereby improving the accuracy of extracted data while maintaining automation.
Solution Approach 2:
The patent introduces image feature values as an intermediary between the voucher image and the extracted data. These image feature values serve as additional information that complements OCR results, enabling the system to verify and correct extracted character strings, thus improving accuracy without sacrificing automation.
2Measurement precision
If human verification is used to check agreement between accounting data and vouchers, then accuracy is maintained, but work efficiency is low
Solution Approach 1:
The patent enables the system to perform self-verification by automatically comparing accounting data with voucher images using extracted image feature values and OCR results. This self-service capability allows the system to verify agreement between accounting data and vouchers without human intervention, thereby improving both accuracy and work efficiency.
Solution Approach 2:
The patent replaces the mechanical human verification process with an automated system that uses image processing and data comparison algorithms. This substitution eliminates manual checking while maintaining verification accuracy, thus significantly improving work efficiency.
3Extent of automation
If OCR is used to extract values from vouchers, then data extraction is automated, but it becomes difficult to determine what the extracted values indicate, requiring human intervention
Solution Approach 1:
The patent segments the voucher image into local regions and extracts image feature values that describe the visual characteristics of different areas. This segmentation provides contextual information about where specific values are located on the voucher, making it easier to automatically determine what the extracted values indicate without human intervention.
Solution Approach 2:
The patent introduces image feature values as an intermediary that bridges the gap between extracted values and their meanings. These image feature values provide contextual information that automatically explains what the extracted values represent, eliminating the need for human intervention in determining value meaning.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
The system effectively checks agreement between vouchers and accounting data, improving efficiency by reducing reliance on human verification and handling incomplete OCR outputs, thus enhancing the accuracy and speed of the voucher verification process.
Implementation Method 1
The text data is data extracted from the image data by optical character recognition
Data Source
AI summary
Provided are a voucher verification method and a voucher verification system for automatically checking agreement between a voucher and accounting data. A processing portion receives accounting data, voucher image data, and text data, extracts a local image feature value from the voucher image data, generates, using a learned determination model, a first vector based on the accounting data and a second vector based on the local image feature value and the text data, calculates similarity between the first vector and the second vector, determines whether the accounting data agrees with the voucher using the calculated similarity, and outputs a determination result. Note that the text data is data extracted from the voucher image data by optical character recognition.