Document Table Reconstruction From Glyph Alignment and Text Blocks
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods struggle to accurately identify and extract tables from unstructured documents like PDF, leading to inefficiencies and frustration during editing due to incomplete or inaccurate representation of table structures.
Innovation Solution
A computer-implemented method that generates document layout data by identifying text segments, blocks, and cells based on alignment, forming editable tables by associating glyphs with document lines and creating links between segments and blocks, thereby reconstructing the table structure.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If automated methods are used to identify tables from unstructured text, then productivity is improved, but measurement precision deteriorates
Solution Approach 1:
The patent segments the unstructured document into discrete text segments and groups them into text blocks based on spatial relationships. Each text block represents a potential table cell, and multiple text blocks are assembled into table structures through link analysis. This segmentation approach enables automated processing while maintaining accuracy by treating each segment independently before reconstruction.
Solution Approach 2:
The patent introduces a new dimension of analysis by creating text block links that represent spatial and structural relationships between text segments. Instead of analyzing text linearly, the system builds a multi-dimensional structure where text blocks are connected through horizontal and vertical relationships, enabling accurate table structure reconstruction from unstructured data.
2Measurement precision
If complex processing is applied to detect table structures, then measurement precision is improved, but device complexity increases
Solution Approach 1:
The system divides the complex task of table detection into manageable segments: text segmentation, text block formation, link creation, and table assembly. Each segment handles a specific aspect of the problem, reducing overall system complexity while maintaining high detection accuracy through systematic processing of individual components.
Solution Approach 2:
The patent introduces text blocks as intermediary structures between raw text segments and final table structures. These text blocks serve as mediators that capture spatial relationships and group related text segments together, simplifying the transition from unstructured text to structured tables without requiring direct complex analysis of all text relationships.
3Measurement precision
If manual table creation is performed, then measurement precision is improved, but loss of time increases
Solution Approach 1:
The system performs preliminary actions by automatically segmenting text, forming text blocks, and creating structural links before final table assembly. This preliminary processing captures the essential structure and relationships early in the workflow, enabling rapid table reconstruction without requiring time-consuming manual analysis while maintaining high accuracy.
Solution Approach 2:
The patent implements self-service by enabling the system to automatically identify table structures, group text segments into appropriate blocks, and assemble tables based on detected relationships without human intervention. The automated process serves itself by using the document's own spatial and structural information to guide the extraction process, eliminating manual time investment while preserving accuracy.
Data Source
AI summary
Embodiments of computer-implemented systems and methods for extracting tables from documents are described. Document layout data is generated based on a plurality of glyphs associated with document lines, comprising data identifying: a plurality of text segments within each line; a plurality of text segment links; a plurality of text blocks of one or more of the text segments; and a plurality of text block links. A document is generated including at least one editable document table corresponding to at least one document table identified based on the plurality of glyphs associated with document lines.