Network graph data extraction method based on attention learning

A technology of data extraction and network diagrams, applied in biological neural network models, instruments, calculations, etc., can solve problems such as matching text and graphics, identifying difficult connecting lines, and high data dimensions, achieving improved robustness and high practical value , the effect of good development prospects

Pending Publication Date: 2022-01-21
EAST CHINA NORMAL UNIVERSITY
View PDF0 Cites 2 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

Studies have shown that there are some methods to solve the problem of obtaining original data, but these methods are only suitable for some simple charts

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Network graph data extraction method based on attention learning
  • Network graph data extraction method based on attention learning
  • Network graph data extraction method based on attention learning

Examples

Experimental program
Comparison scheme
Effect test

Example Embodiment

[0047] Example 1

[0048] Step 1: Enter a think of the size of H × W. Wherein h is the number of image I longitudinal to each column, W is the number of image I horizontally.

[0049] Step 2: Enter the image I input to the text extraction model in step 1, and output the information of each text box t from the output layer of the CRNN neural network after the characteristic of the text information from the CTPN. t . This array contains {t x , T y , T w , T h , T a , Text, Confidence}, where: T x , T y Is the coordinate of the center of this text, T w , T h Is the width and high of this text, T a It is the inclination of the text box. Text is the textual content of the text box. Confidence is the confidence of this text box, and the value of the default confidence is above 0.95 is trusted.

[0050] Step 3: Remove the text box portion in the image I, fill the patch of kernel = (2, 2) with the background color block of the text box.

[0051] Step 4: In the semantic segmentation networ...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

PUM

No PUM Login to view more

Abstract

The invention discloses a network graph data extraction method based on attention learning. The method is characterized by comprising the following steps: extracting character data in a graph by adopting a character extraction deep neural network, extracting pixel characteristics of a visual graph of a network graph by adopting a semantic segmentation network having an attention module, and calculating data of a node and a connection relationship, so that a data structure of an original network graph can be recovered. Compared with the prior art, the method has the advantages that: the problem of data mining of advanced visual coding of network graphs is well solved; the problem of connecting line recognition is solved through an attention mechanism, the robustness of a model is improved; data extraction can be carried out on bitmaps of network graph visualization charts in various practical application scenes such as data conversion, visualization design style switching, and intellectual property protection, and the method has high practical value and good development prospects.

Description

technical field [0001] The invention relates to the technical field of graph data extraction, in particular to a method for extracting original data of network graphs based on attention learning. Background technique [0002] Data visualization images can help people get data characteristics faster, and most charts are stored in the form of bitmap images and published on various media. Obtaining their raw data is a complex task, and recovering the raw data of graphs has become an important research because its development directly contributes to the field of human-computer interaction. [0003] Network diagrams refer to relationship diagrams with text information and different styles, including mind maps, modeling diagrams, flowcharts, etc. They are high-level visual codes that are easy for humans to understand but difficult for machines to decode. Studies have shown that there are some methods to solve the problem of obtaining original data, but these methods are only suit...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

Application Information

Patent Timeline
no application Login to view more
IPC IPC(8): G06V10/774G06V30/146G06V30/41G06K9/62G06N3/04
CPCG06N3/045G06F18/214
Inventor 宋思程王长波李晨辉
Owner EAST CHINA NORMAL UNIVERSITY
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap