PDF Text Block Segmentation Using Gap and Baseline Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
PDF files lack direct information on section and column breaks, making it difficult to accurately transform them into documents with identical segmentations, such as Office Software formats.
Innovation Solution
A method and device that perform an initial division based on character gap outliers, correct errors in text blocks, and compare lines to merge or divide them accurately, ensuring precise segmentation and transformation into target documents.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If PDF text is transformed to another document format, then the document can be edited in Office Software, but the segmentation information (section breaks and column breaks) is lost
Solution Approach 1:
The patent performs preliminary parsing of the PDF file to extract character position information, baseline data, and gap measurements before transformation. This preliminary action captures segmentation cues that would otherwise be lost during format conversion, enabling accurate reconstruction of section and column breaks in the target document format
Solution Approach 2:
The patent introduces an intermediary processing layer that analyzes character gaps, baseline distances, and positional relationships to infer segmentation structure. This intermediary analysis acts as a mediator between the source PDF and target format, translating implicit spatial relationships into explicit segmentation information that preserves document structure during transformation
2Productivity
If character gap outliers are used for initial division, then text blocks can be segmented, but error characters and inaccurate lines are introduced
Solution Approach 1:
The patent implements a feedback mechanism where initially segmented text blocks are evaluated for accuracy by analyzing character positions, baseline consistency, and gap patterns. Inaccurate segments are identified and corrected through iterative refinement, using feedback from the analysis of segmentation quality metrics to improve subsequent segmentation decisions
Solution Approach 2:
The patent performs an initial overly-segmented division using character gap outliers, then selectively merges adjacent blocks that should belong together based on baseline analysis and gap pattern recognition. This approach of excessive initial segmentation followed by selective merging allows rapid initial processing while maintaining the ability to refine accuracy in subsequent steps
3Manufacturing precision
If multiple comparison steps are performed to confirm merging and division, then segmentation accuracy is improved, but processing time increases
Solution Approach 1:
The patent segments the comparison process into distinct hierarchical levels: first comparing characters within lines, then lines within text blocks, and finally text blocks within sections. This multi-level segmentation of the comparison process allows efficient processing at each level while maintaining overall segmentation accuracy, avoiding the need for exhaustive pairwise comparisons of all elements
Data Source
AI summary
There is provided a method for determining text blocks of a PDF text, including: acquiring information of characters of the PDF text; performing an initial division according to gap outliers of the PDF text in a transverse direction and a longitudinal direction, and adding block tags of first text blocks to the information of characters; sequentially processing inaccurate lines and inaccurate words in each first text block according to baselines of characters, character lengths, character spaces and character indexes; performing a baseline arrangement on lines of the PDF text; sequentially comparing two lines to form second text blocks; and sequentially comparing two second text blocks to identify whether to perform a secondary merging and a secondary division.