Table Recognition Using Separator-Based Cell-Text Matching
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing table recognition methods fail to accurately match the location information of cells with their corresponding texts due to insufficient accuracy, leading to incorrect table extraction and degraded user experience.
Innovation Solution
A comprehensive table recognition method that involves obtaining and processing a target image using a neural network model to separately identify cells and separators, performing text recognition, and calculating location information of both cells and separators to enhance matching accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If only model-generated location information of cells is considered for matching, then the processing is simple and fast, but the matching accuracy is insufficient leading to incorrect table extraction
Solution Approach 1:
The patent segments the location information determination into two independent parts: model-generated location information and calculated location information based on separator positions. This segmentation allows each part to be optimized separately while combining their advantages to achieve both accuracy and reasonable complexity
Solution Approach 2:
The patent performs preliminary calculation of cell locations based on separator positions before final matching. By pre-calculating expected cell locations from separator coordinates, the system establishes a reference framework that improves matching accuracy while maintaining processing efficiency
2Reliability
If comprehensive processing of multiple location information sources is performed, then matching accuracy improves, but the processing time and computational cost increase
Solution Approach 1:
The patent introduces separator position information as an intermediary element that mediates between image processing and cell-text matching. By using separator coordinates to calculate expected cell locations, the system creates a reliable reference that accelerates the matching process while ensuring accuracy
Solution Approach 2:
The patent changes the parameter representation by introducing a second location information system based on separator coordinates. This parameter transformation allows the system to verify and correct cell locations through geometric relationships, improving reliability without requiring complex reprocessing
Data Source
Figure 1
Figure 2a
Figure 2b
AI summary
This application discloses a table recognition method and a related device therefor, to correctly complete matching between a plurality of cells and a plurality of texts, so as to obtain a correct table. The method in this application includes: When table recognition needs to be performed, a target image including a to-be-recognized table may be first obtained. After the target image is obtained, the target image may be processed by using a target model, to separately obtain a plurality of cells of the table, first location information of the plurality of cells, and second location information of a plurality of separators between the plurality of cells. After the target image is obtained, text recognition may be further performed on the target image, to obtain a plurality of texts of the table and third location information of the plurality of texts. Then, fourth location information of the plurality of cells may be determined based on the second location information of the plurality of separators. Finally, the location information may be comprehensively considered, to complete matching between the plurality of cells and the plurality of texts, so as to obtain the required table.