Dual Binary Image Layout Analysis for Color Text Extraction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing layout analysis techniques for color images are less accurate, particularly in extracting text blocks from color images, due to limitations in binarization methods and handling multiple background colors and white characters, leading to incomplete extraction of text blocks when characters and graphics are closely arranged.
Innovation Solution
A layout analysis program that employs a dual binary image layout analysis process, using a first binarization process for extracting character elements and a second binarization process for non-character elements, with additional steps to reverse luminance, extract linked black pixels, and determine text blocks based on periodicity of black pixel distribution, ensuring accurate separation and extraction of text blocks.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If a single binarization method is used for color images, then the processing is simple and fast, but the text block extraction accuracy deteriorates, especially for white characters on colored backgrounds and characters near graphics
Solution Approach 1:
The patent divides the layout analysis into two separate binary image analysis processes: one dedicated to character element extraction and another to non-character element extraction. This segmentation allows each process to be optimized for its specific task, with the character-focused process accurately extracting text blocks even from colored backgrounds, while the graphic-focused process handles illustrations and layouts.
Solution Approach 2:
The patent applies different analysis approaches to different regions of the image. By performing character element extraction first and then using that information to guide non-character element extraction, the system applies appropriate processing quality to different parts of the image, ensuring high accuracy for text regions while efficiently handling graphic regions.
2Device complexity
If traditional binary image layout analysis is applied directly to color images, then the process is simple, but the accuracy deteriorates due to inability to handle multiple background colors and white characters
Solution Approach 1:
The patent segments the layout analysis into two distinct binary image analysis processes. The first process extracts character elements with high accuracy by focusing on text-like patterns, while the second process extracts non-character elements by excluding previously identified character regions. This segmentation enables accurate handling of multiple background colors and white characters.
Solution Approach 2:
The patent performs character element extraction as a preliminary step before non-character element extraction. By first identifying and removing character elements from the image, the subsequent non-character element extraction can focus solely on graphics and layouts, significantly improving accuracy for both types of elements.
3Productivity
If character elements and non-character elements are extracted simultaneously, then the processing is efficient, but the separation accuracy deteriorates when characters and graphics are closely arranged
Solution Approach 1:
The patent segments the extraction process into sequential steps: first extracting character elements, then extracting non-character elements from the remaining regions. This segmentation prevents the confusion that occurs when attempting simultaneous extraction of closely arranged characters and graphics, as each extraction process has clear boundaries and focus.
Solution Approach 2:
The patent performs character element extraction as a preliminary action before non-character element extraction. This preliminary extraction removes character elements from consideration, allowing the subsequent non-character element extraction to accurately identify graphics and layouts even when they are closely arranged with text.
Data Source
AI summary
A layout analysis program, a layout analysis apparatus, layout analysis method and a medium can highly accurately extract a text block from an image if the image is a color image. The layout analysis program causes a computer to execute a divided region extracting step that extracts a region partitioned by a pattern according to a binary image so as to use the outcome of extraction as divided region, a set of character elements extracting step that extracts a set of the character elements extracted by a first binary image layout analysis process for each extracted divided region so as to use the outcome of extraction as set of character elements, a text block extracting step that extracts a region including the extracted set of character elements in each divided region so as to avoid overlapping the non-character elements extracted by a second binary image layout analysis process and use the outcome of extraction as text block and a layout information generating step that generates layout information according to the text block and the non-character elements extracted by the second binary image layout analysis process.