Detection and recognition of text superimposed in images base on layered method
Identifieur interne : 000131 ( Main/Exploration ); précédent : 000130; suivant : 000132Detection and recognition of text superimposed in images base on layered method
Auteurs : JIANQIANG YAN [République populaire de Chine] ; XINBO GAO [République populaire de Chine]Source :
- Neurocomputing : (Amsterdam) [ 0925-2312 ] ; 2014.
Descripteurs français
- Pascal (Inist)
- Recherche information, Texte, Reconnaissance caractère, Reconnaissance forme, Reconnaissance optique caractère, Image couleur, Reconnaissance image, Analyse image, Vision ordinateur, Image multiple, Classification automatique, Analyse donnée, Apprentissage supervisé, Analyse amas, Graphe connexe, Modèle agrégé, Résultat expérimental, Localisation.
English descriptors
- KwdEn :
- Aggregate model, Automatic classification, Character recognition, Cluster analysis, Color image, Computer vision, Connected graph, Data analysis, Experimental result, Image analysis, Image recognition, Information retrieval, Localization, Multiple image, Optical character recognition, Pattern recognition, Supervised learning, Text.
Abstract
Detection and recognition of text superimposed in complex background has been considered as a challenging problem. Most of the existing methods first locate the text regions and then feed them into OCR package for recognition. However, these methods cannot achieve good recognition performance due to the complex background. For this purpose, this paper proposes a novel text detection and recognition method by using color clustering to divide images into multiple layers according to main color class. In the proposed method, we exploited a connected component analysis to obtain the candidate text regions from each color layer, and then a cascade Adaboost classifier is adopted to determine whether the candidate text regions is real text regions in the corresponding image layer. Because the monochrome color exists in each layer, the interference of the background can be effectively reduced, which can significantly improve the accuracy of text regions localization. Afterwards, an OCR package is used to recognize the text regions which have been located by the cascade Adaboost classifier. Since the text region has a monochrome color, it helps to greatly improve the recognition rate. Finally, the relationship between different layers is used to verify the recognition results by the text location. The experimental results show that the proposed approach significantly outperforms the existing methods.
Affiliations:
Links toward previous steps (curation, corpus...)
- to stream PascalFrancis, to step Corpus: 000019
- to stream PascalFrancis, to step Curation: 000745
- to stream PascalFrancis, to step Checkpoint: 000014
- to stream Main, to step Merge: 000132
- to stream Main, to step Curation: 000131
Le document en format XML
<record><TEI><teiHeader><fileDesc><titleStmt><title xml:lang="en" level="a">Detection and recognition of text superimposed in images base on layered method</title>
<author><name sortKey="Jianqiang Yan" sort="Jianqiang Yan" uniqKey="Jianqiang Yan" last="Jianqiang Yan">JIANQIANG YAN</name>
<affiliation wicri:level="1"><inist:fA14 i1="01"><s1>School of Electronic Engineering, Xidian University</s1>
<s2>Xi'an 710071</s2>
<s3>CHN</s3>
<sZ>1 aut.</sZ>
<sZ>2 aut.</sZ>
</inist:fA14>
<country>République populaire de Chine</country>
<wicri:noRegion>Xi'an 710071</wicri:noRegion>
</affiliation>
</author>
<author><name sortKey="Xinbo Gao" sort="Xinbo Gao" uniqKey="Xinbo Gao" last="Xinbo Gao">XINBO GAO</name>
<affiliation wicri:level="1"><inist:fA14 i1="01"><s1>School of Electronic Engineering, Xidian University</s1>
<s2>Xi'an 710071</s2>
<s3>CHN</s3>
<sZ>1 aut.</sZ>
<sZ>2 aut.</sZ>
</inist:fA14>
<country>République populaire de Chine</country>
<wicri:noRegion>Xi'an 710071</wicri:noRegion>
</affiliation>
</author>
</titleStmt>
<publicationStmt><idno type="wicri:source">INIST</idno>
<idno type="inist">14-0131436</idno>
<date when="2014">2014</date>
<idno type="stanalyst">PASCAL 14-0131436 INIST</idno>
<idno type="RBID">Pascal:14-0131436</idno>
<idno type="wicri:Area/PascalFrancis/Corpus">000019</idno>
<idno type="wicri:Area/PascalFrancis/Curation">000745</idno>
<idno type="wicri:Area/PascalFrancis/Checkpoint">000014</idno>
<idno type="wicri:doubleKey">0925-2312:2014:Jianqiang Yan:detection:and:recognition</idno>
<idno type="wicri:Area/Main/Merge">000132</idno>
<idno type="wicri:Area/Main/Curation">000131</idno>
<idno type="wicri:Area/Main/Exploration">000131</idno>
</publicationStmt>
<sourceDesc><biblStruct><analytic><title xml:lang="en" level="a">Detection and recognition of text superimposed in images base on layered method</title>
<author><name sortKey="Jianqiang Yan" sort="Jianqiang Yan" uniqKey="Jianqiang Yan" last="Jianqiang Yan">JIANQIANG YAN</name>
<affiliation wicri:level="1"><inist:fA14 i1="01"><s1>School of Electronic Engineering, Xidian University</s1>
<s2>Xi'an 710071</s2>
<s3>CHN</s3>
<sZ>1 aut.</sZ>
<sZ>2 aut.</sZ>
</inist:fA14>
<country>République populaire de Chine</country>
<wicri:noRegion>Xi'an 710071</wicri:noRegion>
</affiliation>
</author>
<author><name sortKey="Xinbo Gao" sort="Xinbo Gao" uniqKey="Xinbo Gao" last="Xinbo Gao">XINBO GAO</name>
<affiliation wicri:level="1"><inist:fA14 i1="01"><s1>School of Electronic Engineering, Xidian University</s1>
<s2>Xi'an 710071</s2>
<s3>CHN</s3>
<sZ>1 aut.</sZ>
<sZ>2 aut.</sZ>
</inist:fA14>
<country>République populaire de Chine</country>
<wicri:noRegion>Xi'an 710071</wicri:noRegion>
</affiliation>
</author>
</analytic>
<series><title level="j" type="main">Neurocomputing : (Amsterdam)</title>
<title level="j" type="abbreviated">Neurocomputing : (Amst.)</title>
<idno type="ISSN">0925-2312</idno>
<imprint><date when="2014">2014</date>
</imprint>
</series>
</biblStruct>
</sourceDesc>
<seriesStmt><title level="j" type="main">Neurocomputing : (Amsterdam)</title>
<title level="j" type="abbreviated">Neurocomputing : (Amst.)</title>
<idno type="ISSN">0925-2312</idno>
</seriesStmt>
</fileDesc>
<profileDesc><textClass><keywords scheme="KwdEn" xml:lang="en"><term>Aggregate model</term>
<term>Automatic classification</term>
<term>Character recognition</term>
<term>Cluster analysis</term>
<term>Color image</term>
<term>Computer vision</term>
<term>Connected graph</term>
<term>Data analysis</term>
<term>Experimental result</term>
<term>Image analysis</term>
<term>Image recognition</term>
<term>Information retrieval</term>
<term>Localization</term>
<term>Multiple image</term>
<term>Optical character recognition</term>
<term>Pattern recognition</term>
<term>Supervised learning</term>
<term>Text</term>
</keywords>
<keywords scheme="Pascal" xml:lang="fr"><term>Recherche information</term>
<term>Texte</term>
<term>Reconnaissance caractère</term>
<term>Reconnaissance forme</term>
<term>Reconnaissance optique caractère</term>
<term>Image couleur</term>
<term>Reconnaissance image</term>
<term>Analyse image</term>
<term>Vision ordinateur</term>
<term>Image multiple</term>
<term>Classification automatique</term>
<term>Analyse donnée</term>
<term>Apprentissage supervisé</term>
<term>Analyse amas</term>
<term>Graphe connexe</term>
<term>Modèle agrégé</term>
<term>Résultat expérimental</term>
<term>Localisation</term>
</keywords>
</textClass>
</profileDesc>
</teiHeader>
<front><div type="abstract" xml:lang="en">Detection and recognition of text superimposed in complex background has been considered as a challenging problem. Most of the existing methods first locate the text regions and then feed them into OCR package for recognition. However, these methods cannot achieve good recognition performance due to the complex background. For this purpose, this paper proposes a novel text detection and recognition method by using color clustering to divide images into multiple layers according to main color class. In the proposed method, we exploited a connected component analysis to obtain the candidate text regions from each color layer, and then a cascade Adaboost classifier is adopted to determine whether the candidate text regions is real text regions in the corresponding image layer. Because the monochrome color exists in each layer, the interference of the background can be effectively reduced, which can significantly improve the accuracy of text regions localization. Afterwards, an OCR package is used to recognize the text regions which have been located by the cascade Adaboost classifier. Since the text region has a monochrome color, it helps to greatly improve the recognition rate. Finally, the relationship between different layers is used to verify the recognition results by the text location. The experimental results show that the proposed approach significantly outperforms the existing methods.</div>
</front>
</TEI>
<affiliations><list><country><li>République populaire de Chine</li>
</country>
</list>
<tree><country name="République populaire de Chine"><noRegion><name sortKey="Jianqiang Yan" sort="Jianqiang Yan" uniqKey="Jianqiang Yan" last="Jianqiang Yan">JIANQIANG YAN</name>
</noRegion>
<name sortKey="Xinbo Gao" sort="Xinbo Gao" uniqKey="Xinbo Gao" last="Xinbo Gao">XINBO GAO</name>
</country>
</tree>
</affiliations>
</record>
Pour manipuler ce document sous Unix (Dilib)
EXPLOR_STEP=$WICRI_ROOT/Ticri/CIDE/explor/OcrV1/Data/Main/Exploration
HfdSelect -h $EXPLOR_STEP/biblio.hfd -nk 000131 | SxmlIndent | more
Ou
HfdSelect -h $EXPLOR_AREA/Data/Main/Exploration/biblio.hfd -nk 000131 | SxmlIndent | more
Pour mettre un lien sur cette page dans le réseau Wicri
{{Explor lien |wiki= Ticri/CIDE |area= OcrV1 |flux= Main |étape= Exploration |type= RBID |clé= Pascal:14-0131436 |texte= Detection and recognition of text superimposed in images base on layered method }}
This area was generated with Dilib version V0.6.32. |