Unlock instant, AI-driven research and patent intelligence for your innovation.

A table extraction method and system based on ellipse fitting in natural scenes

An ellipse fitting, natural scene technology, applied in computing, character and pattern recognition, computer components and other directions, can solve problems such as instability, discontinuous borders, and quadrilaterals that cannot handle perspective transformation, to avoid errors and solve borders. discontinuous effect

Active Publication Date: 2022-05-03
SHANDONG INSPUR SCI RES INST CO LTD
View PDF4 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0005] Aiming at the defects existing in the existing table extraction in natural scenes, the present invention provides a table extraction method and system based on ellipse fitting in natural scenes to deal with perspective-transformed quadrilaterals that cannot be processed by the minimum generated rectangle method, and solve the problem that edge detection methods cannot handle. Deal with the problem of discontinuous borders, solve the problem that straight line detection requires complex post-processing and the effect is unstable

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
  • A table extraction method and system based on ellipse fitting in natural scenes
  • A table extraction method and system based on ellipse fitting in natural scenes

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0048] combined with figure 1 , this embodiment proposes a table extraction method based on ellipse fitting in a natural scene, and the implementation process of the table extraction method includes:

[0049] S10: Obtain an image containing a target table;

[0050] S20: performing adaptive binarization on the acquired image;

[0051] S30: use the findContours function of OpenCV to extract the table frame in the image;

[0052] S40: Perform ellipse fitting on the extracted form frame, and then expand the ellipse frame to obtain four intersection point coordinates;

[0053] S50: By defining the coordinates of the correction template, calculating the projection matrix and performing homography transformation to obtain a rectified rectangular form, which is the finally extracted form.

[0054] In step S10, the image containing the target form is a photo or a plane scanned image without obvious distortion;

[0055] The object form contained on the photo or the image is skewed o...

Embodiment 2

[0064] combined with figure 2 , this embodiment proposes a table extraction system based on ellipse fitting in a natural scene, the table extraction system includes:

[0065] Obtaining module 1, configured to obtain an image containing a target table;

[0066] The binary processing module 2 is used to perform adaptive binarization on the acquired image;

[0067] Extraction module 3, for using the findContours function of OpenCV to extract the table frame in the image;

[0068] The ellipse fitting module 4 is used to perform ellipse fitting on the extracted form frame, and expand the ellipse frame to obtain four intersection point coordinates;

[0069] The calculation and correction module 5 is used to calculate the projection matrix and perform homography transformation by defining the coordinates of the correction template, and then correct the rectangular form, and the corrected rectangular form is the final extracted form.

[0070] In this embodiment, the image containi...

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 table extraction method based on ellipse fitting in a natural scene, which relates to the technical field of table extraction, adopts a technical solution to complete the table frame extraction by using the findContours function of OpenCV on the image after adaptive binarization processing, and further Perform ellipse fitting and expansion on the extracted frame, calculate the projection matrix after obtaining the four corner points, and perform homography transformation to obtain a rectified rectangular table, which is the table extracted from the image. This method weakens the interference of small-scale occlusion and abnormal text, stamps, etc. on the continuity and regularity of the frame by fitting a large number of points. The present invention also provides a form extraction system based on ellipse fitting in a natural scene, which uses the four corner points of the quadrilateral frame and the intersection points of the ellipse to complete the operation of extracting the form from the image, which can avoid errors caused by edge discontinuity and help Straightening and cutting after form extraction.

Description

technical field [0001] The invention relates to the technical field of table extraction, in particular to a table extraction method and system based on ellipse fitting in natural scenes. Background technique [0002] The commonly used form detection and extraction technology is often bound to the Gaopai instrument. Thanks to the dark background of its base, the form extraction and detection speed is faster and the result is more robust. However, the use of equipment for form extraction has certain limitations in real-world scenarios, that is, people cannot perform form detection and extraction without the equipment, and the equipment is expensive and difficult to carry. [0003] Table extraction in natural scenarios can be used in a wider range of scenarios, such as invoice information entry in the financial system, report information extraction, etc. The data sources of these application scenarios are usually only photos and scanned images, so there is no way to guarantee ...

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 Patents(China)
IPC IPC(8): G06V30/412
CPCG06V30/412
Inventor 段强李锐于治楼
Owner SHANDONG INSPUR SCI RES INST CO LTD