OCR Mis-Converted Character Correction via Font Matching
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Image processing systems face challenges in accurately detecting and correcting mis-converted characters in text data obtained through OCR processing, especially when original documents are overlapped or partially missing, leading to incomplete text data.
Innovation Solution
An image processing device that identifies character regions, extracts text, determines matching words from a registered dictionary, and generates corrected text by replacing non-matching characters with font characters having a high degree of matching, using a processor and storage device to analyze and correct mis-converted characters.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If OCR processing is performed on document images with overlapping or partially missing originals, then text data acquisition speed is improved, but text data accuracy deteriorates due to mis-converted characters
Solution Approach 1:
The system extracts text from the document image using OCR, then compares the extracted text against a dictionary to identify non-matching words. For each non-matching word, it analyzes the character image and compares it with font character images to detect mis-converted characters. This feedback loop allows the system to correct accuracy issues while maintaining high processing speed through automated comparison and replacement of mis-converted characters with correct ones from the font database.
Solution Approach 2:
The system introduces an intermediary correction process between OCR extraction and final text output. When a non-matching word is detected, the system uses character image comparison with font characters as an intermediary step to identify the correct character. This intermediary correction mechanism resolves accuracy issues without requiring manual intervention, thereby maintaining productivity while improving text data accuracy.
2Loss of time
If text extraction is performed on incomplete document images, then processing time is reduced, but text completeness deteriorates
Solution Approach 1:
The system implements a feedback mechanism where extracted text is validated against a dictionary. When non-matching words are detected, the system performs character-level analysis and comparison with font characters to identify and correct mis-converted characters. This feedback loop ensures that even when processing incomplete or degraded document images, the system can recover and correct text data, reducing information loss while maintaining efficient processing times through automated correction.
3Measurement precision
If character-level comparison with font characters is performed for correction, then correction accuracy is improved, but processing complexity increases
Solution Approach 1:
The system segments the correction process into distinct stages: first extracting text with OCR, then identifying non-matching words through dictionary comparison,接着 analyzing character images for mis-converted characters, and finally replacing incorrect characters with correct ones from the font database. This segmentation allows the system to achieve high correction accuracy through systematic character-level comparison while managing processing complexity through modular, staged operations rather than attempting all corrections simultaneously.
Data Source
AI summary
An image processing device includes a storage device that previously stores a document image, a plurality of registered words, and a plurality of font characters, and a control device that functions as: a character region identifier that identifies a character region in the document image; an image acquirer that acquires an image of the character region; a text extractor that extracts a text from the image of the character region; a word identifier that identifies each of words in the text; a word determiner that determines whether each of the words is matched with one of the registered words; and a generator that generates a corrected text by replacing a target character of a non-matching word in the text with, among the font characters, a font character having a first degree of matching not lower than a first rate with the target character and a highest first degree of matching.