MSER Fast Skewed Scene-text Location Algorithm

Aiming at the problem that text localization requires a large number of training samples in natural scenes, which leads to low speed of algorithm running and it is difficult to locate skewed text, a fast natural scene skewed text localization algorithm based on maximally stable extremal regions (MSE...

Full description

Saved in:
Bibliographic Details
Main Authors: ZHANG Kai-yu, SHAO Kang-yi, LU Di
Format: Article
Language:zho
Published: Harbin University of Science and Technology Publications 2019-04-01
Series:Journal of Harbin University of Science and Technology
Subjects:
Online Access:https://hlgxb.hrbust.edu.cn/#/digest?ArticleID=1663
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849330124844957696
author ZHANG Kai-yu
SHAO Kang-yi
LU Di
author_facet ZHANG Kai-yu
SHAO Kang-yi
LU Di
author_sort ZHANG Kai-yu
collection DOAJ
description Aiming at the problem that text localization requires a large number of training samples in natural scenes, which leads to low speed of algorithm running and it is difficult to locate skewed text, a fast natural scene skewed text localization algorithm based on maximally stable extremal regions (MSER) with hierarchical clustering is proposed . The method of MSER ellipse fitting is used to select the maximally stable extremal regions of the images, and according to the characteristics of the fitting ellipse and its position on the images, the majority of non-text regions are filtered out and the text candidate regions are selected. By using the idea of hierarchical clustering, the text regions can be clustered gradually and merged into text regions rapidly. Finally the individual text regions are merged into word regions, which can achieve efficient localization of skewed scenes. Experimental results show that the speed of this algorithm has improved significantly without loss of locating accuracy compared with traditional positioning algorithms.
format Article
id doaj-art-652027a6a2cc476bab6b51bedb9d2c42
institution Kabale University
issn 1007-2683
language zho
publishDate 2019-04-01
publisher Harbin University of Science and Technology Publications
record_format Article
series Journal of Harbin University of Science and Technology
spelling doaj-art-652027a6a2cc476bab6b51bedb9d2c422025-08-20T03:47:03ZzhoHarbin University of Science and Technology PublicationsJournal of Harbin University of Science and Technology1007-26832019-04-012402818810.15938/j.jhust.2019.02.012MSER Fast Skewed Scene-text Location AlgorithmZHANG Kai-yu0SHAO Kang-yi1LU Di2School of Electrical and Electronic Engineering,Harbin University of Science and Technology, Harbin 150080, ChinaSchool of Electrical and Electronic Engineering,Harbin University of Science and Technology, Harbin 150080, ChinaSchool of Electrical and Electronic Engineering,Harbin University of Science and Technology, Harbin 150080, ChinaAiming at the problem that text localization requires a large number of training samples in natural scenes, which leads to low speed of algorithm running and it is difficult to locate skewed text, a fast natural scene skewed text localization algorithm based on maximally stable extremal regions (MSER) with hierarchical clustering is proposed . The method of MSER ellipse fitting is used to select the maximally stable extremal regions of the images, and according to the characteristics of the fitting ellipse and its position on the images, the majority of non-text regions are filtered out and the text candidate regions are selected. By using the idea of hierarchical clustering, the text regions can be clustered gradually and merged into text regions rapidly. Finally the individual text regions are merged into word regions, which can achieve efficient localization of skewed scenes. Experimental results show that the speed of this algorithm has improved significantly without loss of locating accuracy compared with traditional positioning algorithms.https://hlgxb.hrbust.edu.cn/#/digest?ArticleID=1663scene textmaximally stable extremal regionshierarchical clusteringellipse fitting
spellingShingle ZHANG Kai-yu
SHAO Kang-yi
LU Di
MSER Fast Skewed Scene-text Location Algorithm
Journal of Harbin University of Science and Technology
scene text
maximally stable extremal regions
hierarchical clustering
ellipse fitting
title MSER Fast Skewed Scene-text Location Algorithm
title_full MSER Fast Skewed Scene-text Location Algorithm
title_fullStr MSER Fast Skewed Scene-text Location Algorithm
title_full_unstemmed MSER Fast Skewed Scene-text Location Algorithm
title_short MSER Fast Skewed Scene-text Location Algorithm
title_sort mser fast skewed scene text location algorithm
topic scene text
maximally stable extremal regions
hierarchical clustering
ellipse fitting
url https://hlgxb.hrbust.edu.cn/#/digest?ArticleID=1663
work_keys_str_mv AT zhangkaiyu mserfastskewedscenetextlocationalgorithm
AT shaokangyi mserfastskewedscenetextlocationalgorithm
AT ludi mserfastskewedscenetextlocationalgorithm