Document Image Processing System with Handwritten Recognition Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The existing image processing systems face difficulties in efficiently handling and registering both handwritten and printed documents in expense processing systems, leading to cumbersome operations due to lower recognition accuracy of handwritten characters by OCR and the need for multiple scanning processes.
Innovation Solution
An image processing system that includes a generation unit for scanning documents, a controlling unit to associate handwritten document image data with specific information, and a transmission unit to transmit image data to an information processing apparatus, where character recognition processing is executed differently for handwritten and printed documents, allowing for visual checking of handwritten documents and automatic OCR for printed documents within a single scanning process.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If character recognition processing is executed on both handwritten and printed documents using the same scanning process, then the scanning process can be simplified, but the recognition accuracy for handwritten characters deteriorates
Solution Approach 1:
The patent segments the document processing into two distinct paths: handwritten documents are processed through visual checking by operators, while printed documents are processed through automatic OCR character recognition. This segmentation allows each document type to receive appropriate processing methods, maintaining both operational simplicity and recognition accuracy.
Solution Approach 2:
The patent applies different processing qualities to different document types. Handwritten documents receive manual visual checking quality, while printed documents receive automatic OCR quality. This local differentiation ensures that each document type receives the appropriate level of processing precision without compromising overall system efficiency.
2Measurement precision
If multiple scanning processes are executed to separate handwritten and printed documents, then the recognition accuracy for handwritten characters is maintained, but the operator workload increases
Solution Approach 1:
The patent performs preliminary classification of documents into handwritten and printed categories during the scanning process. This preliminary action enables the system to automatically route documents to the appropriate processing path, maintaining recognition accuracy while eliminating the need for operators to manually separate documents through multiple scanning processes.
Solution Approach 2:
The system performs self-classification of document types during scanning, automatically identifying whether a document is handwritten or printed without requiring operator intervention. This self-service capability maintains processing accuracy while significantly improving operator efficiency by eliminating manual document separation tasks.
3Speed
If automatic OCR processing is applied to all documents, then the processing speed increases, but the reliability of character recognition for handwritten documents deteriorates
Solution Approach 1:
The patent implements a dynamic processing approach where the system automatically adapts its processing method based on document type. For printed documents, it applies fast automatic OCR processing; for handwritten documents, it switches to reliable manual visual checking. This dynamic adjustment maintains both processing speed and recognition reliability by selecting the appropriate method for each document type.
Solution Approach 2:
The patent changes the processing parameters based on document characteristics. The system detects whether a document is handwritten or printed and adjusts the processing method accordingly - using OCR for printed documents and visual checking for handwritten documents. This parameter change ensures that processing speed and reliability are optimized for each document type.
Data Source
AI summary
An image processing system includes an information processing apparatus and an image processing apparatus. The image processing apparatus generates a plurality of pieces of image data by scanning a plurality of documents, associates, with image data generated by scanning a handwritten document out of the plurality of pieces of image data, information indicating that the image data is image data generated by scanning a handwritten document, and transmits, to the information processing apparatus, the plurality of pieces of image data and the information. The information processing apparatus receives the information and the plurality of pieces of image data; and executes character recognition processing on first image data with which the information is not associated, and obtains character information based on the first image data, wherein character information based on second image data with which the information is associated is obtained by executing processing different from the character recognition processing.