Table structure identification method, system and equipment based on cell detection

A form structure and recognition method technology, applied in the field of image recognition, to achieve the effect of simple steps, strong universality, and expanded application scenarios

Pending Publication Date: 2022-04-15
SOUTH CHINA UNIV OF TECH
View PDF0 Cites 4 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0004] The form recognition task has been proposed for more than 20 years. At the beginning, it mostly processed documents in pdf format and scanned image data. However, with the development of technology and the widespread popularization of smart devices such as mobile phones, people will use mobile phones to obtain form information. Convenience, but the currently proposed algorithms are less aimed at photographing forms in natural scenes, so in order to promote the development of form recognition technology, it is essential to design a system that is suitable for form structure recognition of photographing forms in natural scenes

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
  • Table structure identification method, system and equipment based on cell detection
  • Table structure identification method, system and equipment based on cell detection
  • Table structure identification method, system and equipment based on cell detection

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0048] Such as figure 1 As shown, this embodiment discloses a method for identifying a table structure based on cell detection, comprising the following steps:

[0049] S1. Obtaining the form area, locating the area where the form is located from the form image through the form area detection model, and generating a new form image.

[0050]The input image in this embodiment is a photographed image of a table in a natural scene, and the form of the image is diverse, including complex illumination changes, background changes, and table shape changes, that is, the table in the image may have various skewed angles or distortions Gesture, the present invention expands the application scenarios of table structure recognition in natural scenarios.

[0051] Preferably, the table region detection model in this example is a table region detection model based on a cascade cyclic convolutional neural network algorithm (CascadeRCNN).

[0052] Specifically, step S1 includes the following ...

Embodiment 2

[0115] The present embodiment provides a table structure recognition system based on cell detection. The system includes a table area acquisition module, a cell detection module, a table structure prediction module, and a table structure visualization recovery module. The specific functions of each module are as follows:

[0116] The form area acquisition module is used to locate the area where the form is located from the form image through the form area detection model to generate a new form image;

[0117]The cell detection module is used to construct a cell detection model based on the improved SBD algorithm by improving the general sequential frame-free decomposition network (SBD) algorithm. The cell detection model detects all cells in the table area and obtains the parcel unit The coordinates of the four vertices of the smallest quadrilateral of the grid;

[0118] The table structure prediction module is used to design a cell adjacency matching algorithm, find cells in ...

Embodiment 3

[0121] This embodiment provides a computer device, which may be a server, a computer, etc., and includes a processor, a memory, an input device, a display, and a network interface connected through a system bus, and the processor is used to provide computing and control capabilities, The memory includes a non-volatile storage medium and an internal memory, the non-volatile storage medium stores an operating system, a computer program and a database, and the internal memory provides for the operation of the operating system and the computer program in the non-volatile storage medium Environment, when the processor executes the computer program stored in the memory, realize the method for a kind of form structure recognition of the above-mentioned embodiment 1, as follows:

[0122] Form area acquisition, locate the area where the form is located from the form image through the form area detection model, and generate a new form image;

[0123] Cell detection, by improving the gen...

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 relates to the field of image recognition, in particular to a table structure recognition method, system and device based on cell detection, and the method comprises the steps: locating a table region from a table image through a table region detection model, and generating a new table image; constructing a cell detection model based on an improved SBD algorithm through the improved universal SBD algorithm, and detecting all cells in the table region by the cell detection model to obtain four vertex coordinates of a minimum quadrangle wrapping the cells; by designing a cell adjacency matching algorithm, finding out cells in the same row or the same column according to the detected coordinates of the cells, and predicting an HTML structure corresponding to the table through a table row and column clustering result; and according to the predicted table HTML structure, restoring and generating a table in an editable form with the same table structure in the image. According to the method, the problem of structure recovery of cross-row and cross-column cells can be solved, and compared with an existing method, the method has higher universality.

Description

technical field [0001] The invention relates to the field of image recognition, in particular to a table structure recognition method, system and equipment based on cell detection. Background technique [0002] Tables are an important way to record and transmit information in our lives. Compared with natural language, tables provide a more compact and structured data format, which can summarize a large amount of data. The form of comparative information is convenient for readers to quickly obtain effective information, so tables are widely used in various fields. Today, with the rapid development of layout analysis and document understanding, tables, as an important part of documents, have great research value. In order to quickly obtain table information, researchers have proposed table recognition tasks. [0003] Table recognition tasks can be decomposed into three parts, table detection, table structure recognition and table text recognition. With the help of today's mat...

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
Patent Type & Authority Applications(China)
IPC IPC(8): G06V30/413G06V30/414G06V10/762G06V10/74G06V10/82G06K9/62G06N3/04G06N3/08
Inventor 薛洋彭帆金连文
Owner SOUTH CHINA UNIV OF TECH
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