Method for quickly locating oblique text in natural scene based on MSER

A natural scene and text technology, applied in the field of locating inclined text, can solve the problems of slow positioning speed, low accuracy, and poor effect

Inactive Publication Date: 2018-07-17
HARBIN UNIV OF SCI & TECH
View PDF7 Cites 6 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

The fitting ellipse text area it adopts can quickly and accurately extract the text information in the image, but the effect is not good when dealing with oblique text, and the accuracy is not high, and a large number of sample training is required, resulting in slow positioning speed

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
  • Method for quickly locating oblique text in natural scene based on MSER
  • Method for quickly locating oblique text in natural scene based on MSER
  • Method for quickly locating oblique text in natural scene based on MSER

Examples

Experimental program
Comparison scheme
Effect test

specific Embodiment approach 1

[0063] A method based on MSER to quickly locate skewed text in natural scenes, such as figure 1 shown, including the following steps:

[0064] Step a, grayscale processing is performed on the original image, the effect is as follows figure 2 shown;

[0065] Step b. Use the MSER algorithm to quickly extract the suspected text area on the grayscale image, and process the suspected text area through ellipse fitting. The processing effect is as follows image 3 shown;

[0066] Step c, according to the ratio of the major and minor axes of the fitted ellipse and the characteristics that the fitted ellipse is centered in the text area, perform coarse filtering of the non-text area; when it does not meet the ratio of the major and minor axes of the fitted ellipse, delete it; when it meets the fitted ellipse When the ratio of the major and minor axes is used, judge whether the fitted ellipse is located at the edge of the image; if so, delete it; if not, proceed to step d;

[0067]...

specific Embodiment approach 2

[0104] The present invention runs on a 64-bit win10 computer with a CPU of i5, a main frequency of 3.30GHZ, and a memory of 8G. The running platform is the public test platform provided by the ICDAR contest, and the used test data set is the ICDAR2015 text positioning data set. 500 test images. The present invention is compared with a number of excellent algorithms of ICDAR2015 and the algorithm of MSER combined with two-way projection. The content of the comparison includes two aspects. One is to use the f value to measure the positioning accuracy of the algorithm, and the other is to use the algorithm to process images in the data set The average time to measure the operation speed of the algorithm.

[0105] Table 1 Comparison of positioning parameters of multiple algorithms

[0106]

[0107] Table 2 Comparison of running time of multiple algorithms

[0108]

[0109] It can be seen from Table 1 that the present invention has certain advantages in terms of recall rate...

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 provides a method for quickly locating an oblique text in a natural scene based on MSER, wherein the method belongs to the technical field of text locating identification. The method comprises the steps of performing gray processing on an original image; performing quick extraction on a suspected text area on a gray image by means of an MSER algorithm, and processing the suspected text area through elliptical processing; according to a long-to-short axis ratio of a fitting ellipse and a characteristic that the fitting ellipse exists at the central part of the text area, performing coarse filtering of a non-text area; according to the ratio between the irregular actual extraction area and the fitting elliptical area of the MSER, performing further fine filtering, and determining whether the fitting ellipse is in an area error; performing text fusion by means of a hierarchical clustering concept, and combining disperse text areas for obtaining a word area through characterspace coarse classifying, character chain member filtering, similarity text chain combining, RGB average value single-character and character chain screening; and obtaining the text area. The method according to the invention has advantages of reducing text location time, realizing no location precision reduction, and realizing high-efficiency oblique scene text locating.

Description

technical field [0001] The invention discloses a method for quickly locating oblique texts in a natural scene based on MSER, which belongs to the technical field of text positioning and recognition. Background technique [0002] The MSER algorithm (maximally stable extremal regions, maximum stable extremal regions) is an affine feature region extraction algorithm. The fitting ellipse text area it adopts can extract the text information in the image relatively quickly and accurately, but the effect is not good when dealing with oblique text, the accuracy is not high, and a large number of sample training is required, resulting in slow positioning speed. Contents of the invention [0003] In view of the above problems, the object of the present invention is to provide a method for quickly locating inclined text in natural scenes based on MSER. [0004] The purpose of the present invention is achieved like this: [0005] A method for quickly locating skewed text in natural ...

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): G06K9/32G06K9/62
CPCG06V20/63G06V30/1478G06F18/241
Inventor 张开玉冯忠明万康怡王强陈荣琪
Owner HARBIN UNIV OF SCI & 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