OCR Page Layout Determination Using White Space Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing optical character recognition (OCR) processes face challenges in accurately determining the page layout and reading order of documents with complex layouts, such as multiple columns and interspersed images, which affects recognition accuracy and retention of the original layout.
Innovation Solution
A method is developed to identify page layouts by grouping words into text regions using whitespace rectangles, sorting them based on vertical position, and determining the reading order through a series of steps involving whitespace candidate seed selection, expansion, and filtering, followed by text region detection and sorting.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional OCR processes are used on documents with complex layouts, then processing speed is maintained, but layout accuracy and reading order determination deteriorate
Solution Approach 1:
The patent segments the document processing task into distinct phases: white space region identification, text region detection, line grouping, and reading order determination. Each phase handles a specific aspect of layout analysis, improving overall accuracy while maintaining manageable complexity through modular processing steps.
Solution Approach 2:
The patent performs preliminary identification of white space regions before detecting text regions. This preliminary action establishes a framework that guides subsequent text region detection and reading order determination, ensuring accurate layout analysis from the outset rather than attempting to determine reading order directly.
2Measurement precision
If simple text region grouping methods are used, then processing speed is maintained, but reading order accuracy deteriorates
Solution Approach 1:
The patent introduces a vertical dimension to reading order determination by analyzing the vertical positions of text regions and lines. Instead of relying solely on horizontal scanning, the method incorporates vertical coordinate analysis to accurately determine reading sequence, especially for complex layouts with multiple columns and interspersed images.
Solution Approach 2:
The patent uses white space regions as an intermediary element to facilitate reading order determination. By identifying and utilizing white space regions as reference points, the system can more accurately determine the spatial relationships between text regions and establish correct reading order without requiring complex direct comparisons of all text elements.
3Measurement precision
If detailed white space analysis is performed, then text region detection accuracy is improved, but computational complexity increases
Solution Approach 1:
The patent changes the parameter being analyzed from direct text region boundaries to white space region characteristics. By analyzing the size, shape, and position of white space regions rather than directly processing text region boundaries, the method achieves accurate text region detection with reduced computational complexity, as white space regions provide clear geometric constraints.
Data Source
Figure 1~2
Figure 3~4
Figure 5~6
AI summary
A method and system is provided for identifying a page layout of an image that includes textual regions. The textual regions are to undergo optical character recognition (OCR). The system includes an input component that receives an input image that includes words around which bounding boxes have been formed and a text identifying component that groups the words into a plurality of text regions. A reading line component groups words within each of the text regions into reading lines. A text region sorting component that sorts the text regions in accordance with their reading order.