Image Processing Apparatus for Document Character Extraction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image processing techniques struggle to accurately extract character information from document images when a color logo mark or imprint overlaps the character portion, as this can interfere with OCR processing.
Innovation Solution
An image processing apparatus that performs binarization processing on a document image to generate a binary image, and then modifies this image to remove background objects overlapping characters by changing specific pixels in the binary image, allowing for accurate separation of characters and background.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If binarization processing is performed on a document image with overlapping logo marks and characters, then the processing is simple and fast, but the character information cannot be extracted accurately due to color interference
Solution Approach 1:
The patent segments the binarization process into multiple stages: first binarization to separate dark characters from light background, then second binarization to separate colored logo marks from characters. This multi-stage segmentation allows each stage to focus on specific separation tasks, resolving the contradiction between simple processing and accurate extraction.
Solution Approach 2:
The patent applies different binarization thresholds and processing methods to different regions of the image. The first binarization uses a threshold suitable for black characters on light background, while the second binarization uses different thresholds to separate colored logo marks. This local differentiation of processing quality enables accurate extraction without compromising overall processing efficiency.
2Device complexity
If the area determined to be a character area is taken as processing target, then the processing scope is limited, but it is not possible to separate characters and background when logo marks overlap characters
Solution Approach 1:
The patent performs preliminary binarization processing to identify and extract the character area before attempting separation. This preliminary action creates a focused processing target that guides subsequent separation operations, enabling accurate character-background separation within the character area while maintaining limited processing scope.
Solution Approach 2:
The patent introduces an intermediate binary image as a mediator between the original document image and the final separated result. This intermediate representation contains only character area information, serving as a bridge that enables precise separation operations while keeping the overall processing scope controlled and manageable.
Data Source
AI summary
To make it possible to extract character information with a high accuracy even from a document image obtained by reading a document in which a logo mark or the like overlaps a character portion. By performing binarization processing for a document image obtained by reading a document, a binary image including first pixels representing a color darker than a reference and second pixels representing a color paler than the reference is generated. Then, by changing the pixel among the first pixels included in the generated binary image, whose corresponding pixel's color in the document image is different from a color of a character object within the document, to the second pixel, a binary image in which a background object that overlaps the character object in the document image is removed is generated.