Image Layout Analysis Using Geometric Paragraph Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing layout analysis methods for recognizing characters in images are computationally heavy and complex, requiring direct image processing and semantic analysis to distinguish between reading material and background characters, which complicates the process.
Innovation Solution
A computer-implemented layout analysis method that divides image paragraphs into columns based on coordinate information, determines main paragraphs using geometric criteria, and classifies non-main paragraphs as additional main paragraphs to refine the layout, thereby reducing computational complexity and eliminating the need for complex image processing or semantic analysis.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If image processing or semantic analysis algorithms are used to distinguish reading material characters from background characters, then layout refinement accuracy is improved, but algorithmic complexity and computational load increase
Solution Approach 1:
The patent segments the image processing task into distinct stages: paragraph detection, column division, main paragraph identification, and additional main paragraph classification. Each stage processes specific geometric features (coordinates, widths, heights, positions) independently, avoiding the need for complex holistic image processing or semantic analysis while maintaining refinement accuracy.
2Measurement precision
If direct image processing is performed to recognize characters in reading material layout, then character recognition accuracy is improved, but computational load increases
Solution Approach 1:
The patent performs preliminary geometric analysis of paragraph structures (detecting coordinates, widths, heights, and positions) before character recognition. By pre-identifying main paragraphs and additional main paragraphs through geometric relationships, the system reduces the scope of subsequent character recognition to only relevant regions, thereby reducing computational load while maintaining accuracy.
3Manufacturing precision
If semantic analysis is used to identify background characters for removal, then layout refinement precision is improved, but processing time increases
Solution Approach 1:
The patent changes the analysis parameters from semantic character content to geometric paragraph features (coordinates, widths, heights, positions, and spatial relationships). By evaluating geometric relationships between paragraphs to identify main and additional main paragraphs, the system achieves layout refinement precision without the time-consuming semantic analysis of character meanings.
Data Source
AI summary
The present application relates to layout analysis on an image. The layout analysis method includes: dividing, based on coordinate information of a plurality of paragraphs in an image, the plurality of paragraphs into one or more columns arranged in a horizontal direction, each column including one or more paragraphs of the plurality of paragraphs; for one or more paragraphs included in each of at least some of the one or more columns, determining a main paragraph in the column based on a first criterion related to geometric information of a paragraph; and for each of the columns, if one or more non-main paragraphs and the main paragraph in the column satisfy a geometric relationship for adding a main paragraph, taking the one or more non-main paragraphs as additional main paragraphs to the main paragraph.