OcrV1, Main, Exploration, bibRecord, 001185

Adaptive pre-OCR cleanup of grayscale document images

Identifieur interne : 001185 ( Main/Exploration ); précédent : 001184; suivant : 001186

Adaptive pre-OCR cleanup of grayscale document images

Auteurs : Ilya Zavorin [États-Unis] ; Eugene Borovikov [États-Unis] ; Mark Turner [États-Unis] ; Luis Hernandez [États-Unis]

Source :

Proceedings of SPIE, the International Society for Optical Engineering [ 0277-786X ] ; 2006.

RBID : Pascal:07-0376469

Descripteurs français

Pascal (Inist)
- Réseau neuronal, Algorithme, Etude expérimentale, Reconnaissance optique caractère, Echelle gris, Traitement image document, Accentuation image, Qualité image, Apprentissage, Détection seuil, Réduction bruit, Implémentation, 0705M.

English descriptors

KwdEn :
- Algorithms, Document image processing, Experimental study, Gray scale, Image enhancement, Image quality, Implementation, Learning, Neural networks, Noise reduction, Optical character recognition, Threshold detection.

Abstract

This paper describes new capabilities of ImageRefiner, an automatic image enhancement system based on machine learning (ML). ImageRefiner was initially designed as a pre-OCR cleanup filter for bitonal (black-and-white) document images. Using a single neural network, ImageRefiner learned which image enhancement transformations (filters) were best suited for a given document image and a given OCR engine, based on various image measurements (characteristics). The new release improves ImageRefiner in three major ways. First, to process grayscale document images, we have included three grayscale filters based on smart thresholding and noise filtering, as well as five image characteristics that are all byproducts of various thresholding techniques. Second, we have implemented additional ML algorithms, including a neural network ensemble and several "all-pairs" classifiers. Third, we have introduced a measure that evaluates overall performance of the system in terms of cumulative improvement of OCR accuracy. Our experiments indicate that OCR accuracy on enhanced grayscale images is higher than that of both the original grayscale images and the corresponding bitonal images obtained by scanning the same documents. We have noticed that the system's performance may suffer when document characteristics are correlated.

Affiliations:

Links toward previous steps (curation, corpus...)

to stream PascalFrancis, to step Corpus: 000337
to stream PascalFrancis, to step Curation: 000449
to stream PascalFrancis, to step Checkpoint: 000356
to stream Main, to step Merge: 001217
to stream Main, to step Curation: 001185

Le document en format XML

<record><TEI><teiHeader><fileDesc><titleStmt><title xml:lang="en" level="a">Adaptive pre-OCR cleanup of grayscale document images</title>
<author><name sortKey="Zavorin, Ilya" sort="Zavorin, Ilya" uniqKey="Zavorin I" first="Ilya" last="Zavorin">Ilya Zavorin</name>
<affiliation wicri:level="2"><inist:fA14 i1="01"><s1>CACI International Inc, 4831 Walden Lane</s1>
<s2>Lanham, MD 20706</s2>
<s3>USA</s3>
<sZ>1 aut.</sZ>
<sZ>2 aut.</sZ>
<sZ>3 aut.</sZ>
</inist:fA14>
<country>États-Unis</country>
<placeName><region type="state">Maryland</region>
</placeName>
</affiliation>
</author>
<author><name sortKey="Borovikov, Eugene" sort="Borovikov, Eugene" uniqKey="Borovikov E" first="Eugene" last="Borovikov">Eugene Borovikov</name>
<affiliation wicri:level="2"><inist:fA14 i1="01"><s1>CACI International Inc, 4831 Walden Lane</s1>
<s2>Lanham, MD 20706</s2>
<s3>USA</s3>
<sZ>1 aut.</sZ>
<sZ>2 aut.</sZ>
<sZ>3 aut.</sZ>
</inist:fA14>
<country>États-Unis</country>
<placeName><region type="state">Maryland</region>
</placeName>
</affiliation>
</author>
<author><name sortKey="Turner, Mark" sort="Turner, Mark" uniqKey="Turner M" first="Mark" last="Turner">Mark Turner</name>
<affiliation wicri:level="2"><inist:fA14 i1="01"><s1>CACI International Inc, 4831 Walden Lane</s1>
<s2>Lanham, MD 20706</s2>
<s3>USA</s3>
<sZ>1 aut.</sZ>
<sZ>2 aut.</sZ>
<sZ>3 aut.</sZ>
</inist:fA14>
<country>États-Unis</country>
<placeName><region type="state">Maryland</region>
</placeName>
</affiliation>
</author>
<author><name sortKey="Hernandez, Luis" sort="Hernandez, Luis" uniqKey="Hernandez L" first="Luis" last="Hernandez">Luis Hernandez</name>
<affiliation wicri:level="2"><inist:fA14 i1="02"><s1>Army Research Laboratory</s1>
<s2>Adelphi, MD</s2>
<s3>USA</s3>
<sZ>4 aut.</sZ>
</inist:fA14>
<country>États-Unis</country>
<placeName><region type="state">Maryland</region>
</placeName>
</affiliation>
</author>
</titleStmt>
<publicationStmt><idno type="wicri:source">INIST</idno>
<idno type="inist">07-0376469</idno>
<date when="2006">2006</date>
<idno type="stanalyst">PASCAL 07-0376469 INIST</idno>
<idno type="RBID">Pascal:07-0376469</idno>
<idno type="wicri:Area/PascalFrancis/Corpus">000337</idno>
<idno type="wicri:Area/PascalFrancis/Curation">000449</idno>
<idno type="wicri:Area/PascalFrancis/Checkpoint">000356</idno>
<idno type="wicri:doubleKey">0277-786X:2006:Zavorin I:adaptive:pre:ocr</idno>
<idno type="wicri:Area/Main/Merge">001217</idno>
<idno type="wicri:Area/Main/Curation">001185</idno>
<idno type="wicri:Area/Main/Exploration">001185</idno>
</publicationStmt>
<sourceDesc><biblStruct><analytic><title xml:lang="en" level="a">Adaptive pre-OCR cleanup of grayscale document images</title>
<author><name sortKey="Zavorin, Ilya" sort="Zavorin, Ilya" uniqKey="Zavorin I" first="Ilya" last="Zavorin">Ilya Zavorin</name>
<affiliation wicri:level="2"><inist:fA14 i1="01"><s1>CACI International Inc, 4831 Walden Lane</s1>
<s2>Lanham, MD 20706</s2>
<s3>USA</s3>
<sZ>1 aut.</sZ>
<sZ>2 aut.</sZ>
<sZ>3 aut.</sZ>
</inist:fA14>
<country>États-Unis</country>
<placeName><region type="state">Maryland</region>
</placeName>
</affiliation>
</author>
<author><name sortKey="Borovikov, Eugene" sort="Borovikov, Eugene" uniqKey="Borovikov E" first="Eugene" last="Borovikov">Eugene Borovikov</name>
<affiliation wicri:level="2"><inist:fA14 i1="01"><s1>CACI International Inc, 4831 Walden Lane</s1>
<s2>Lanham, MD 20706</s2>
<s3>USA</s3>
<sZ>1 aut.</sZ>
<sZ>2 aut.</sZ>
<sZ>3 aut.</sZ>
</inist:fA14>
<country>États-Unis</country>
<placeName><region type="state">Maryland</region>
</placeName>
</affiliation>
</author>
<author><name sortKey="Turner, Mark" sort="Turner, Mark" uniqKey="Turner M" first="Mark" last="Turner">Mark Turner</name>
<affiliation wicri:level="2"><inist:fA14 i1="01"><s1>CACI International Inc, 4831 Walden Lane</s1>
<s2>Lanham, MD 20706</s2>
<s3>USA</s3>
<sZ>1 aut.</sZ>
<sZ>2 aut.</sZ>
<sZ>3 aut.</sZ>
</inist:fA14>
<country>États-Unis</country>
<placeName><region type="state">Maryland</region>
</placeName>
</affiliation>
</author>
<author><name sortKey="Hernandez, Luis" sort="Hernandez, Luis" uniqKey="Hernandez L" first="Luis" last="Hernandez">Luis Hernandez</name>
<affiliation wicri:level="2"><inist:fA14 i1="02"><s1>Army Research Laboratory</s1>
<s2>Adelphi, MD</s2>
<s3>USA</s3>
<sZ>4 aut.</sZ>
</inist:fA14>
<country>États-Unis</country>
<placeName><region type="state">Maryland</region>
</placeName>
</affiliation>
</author>
</analytic>
<series><title level="j" type="main">Proceedings of SPIE, the International Society for Optical Engineering</title>
<idno type="ISSN">0277-786X</idno>
<imprint><date when="2006">2006</date>
</imprint>
</series>
</biblStruct>
</sourceDesc>
<seriesStmt><title level="j" type="main">Proceedings of SPIE, the International Society for Optical Engineering</title>
<idno type="ISSN">0277-786X</idno>
</seriesStmt>
</fileDesc>
<profileDesc><textClass><keywords scheme="KwdEn" xml:lang="en"><term>Algorithms</term>
<term>Document image processing</term>
<term>Experimental study</term>
<term>Gray scale</term>
<term>Image enhancement</term>
<term>Image quality</term>
<term>Implementation</term>
<term>Learning</term>
<term>Neural networks</term>
<term>Noise reduction</term>
<term>Optical character recognition</term>
<term>Threshold detection</term>
</keywords>
<keywords scheme="Pascal" xml:lang="fr"><term>Réseau neuronal</term>
<term>Algorithme</term>
<term>Etude expérimentale</term>
<term>Reconnaissance optique caractère</term>
<term>Echelle gris</term>
<term>Traitement image document</term>
<term>Accentuation image</term>
<term>Qualité image</term>
<term>Apprentissage</term>
<term>Détection seuil</term>
<term>Réduction bruit</term>
<term>Implémentation</term>
<term>0705M</term>
</keywords>
</textClass>
</profileDesc>
</teiHeader>
<front><div type="abstract" xml:lang="en">This paper describes new capabilities of ImageRefiner, an automatic image enhancement system based on machine learning (ML). ImageRefiner was initially designed as a pre-OCR cleanup filter for bitonal (black-and-white) document images. Using a single neural network, ImageRefiner learned which image enhancement transformations (filters) were best suited for a given document image and a given OCR engine, based on various image measurements (characteristics). The new release improves ImageRefiner in three major ways. First, to process grayscale document images, we have included three grayscale filters based on smart thresholding and noise filtering, as well as five image characteristics that are all byproducts of various thresholding techniques. Second, we have implemented additional ML algorithms, including a neural network ensemble and several "all-pairs" classifiers. Third, we have introduced a measure that evaluates overall performance of the system in terms of cumulative improvement of OCR accuracy. Our experiments indicate that OCR accuracy on enhanced grayscale images is higher than that of both the original grayscale images and the corresponding bitonal images obtained by scanning the same documents. We have noticed that the system's performance may suffer when document characteristics are correlated.</div>
</front>
</TEI>
<affiliations><list><country><li>États-Unis</li>
</country>
<region><li>Maryland</li>
</region>
</list>
<tree><country name="États-Unis"><region name="Maryland"><name sortKey="Zavorin, Ilya" sort="Zavorin, Ilya" uniqKey="Zavorin I" first="Ilya" last="Zavorin">Ilya Zavorin</name>
</region>
<name sortKey="Borovikov, Eugene" sort="Borovikov, Eugene" uniqKey="Borovikov E" first="Eugene" last="Borovikov">Eugene Borovikov</name>
<name sortKey="Hernandez, Luis" sort="Hernandez, Luis" uniqKey="Hernandez L" first="Luis" last="Hernandez">Luis Hernandez</name>
<name sortKey="Turner, Mark" sort="Turner, Mark" uniqKey="Turner M" first="Mark" last="Turner">Mark Turner</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 001185 | SxmlIndent | more

HfdSelect -h $EXPLOR_AREA/Data/Main/Exploration/biblio.hfd -nk 001185 | 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:07-0376469
   |texte=   Adaptive pre-OCR cleanup of grayscale document images
}}

This area was generated with Dilib version V0.6.32.
Data generation: Sat Nov 11 16:53:45 2017. Site generation: Mon Mar 11 23:15:16 2024

	Serveur d'exploration sur l'OCR
	Attention, ce site est en cours de développement ! Attention, site généré par des moyens informatiques à partir de corpus bruts. Les informations ne sont donc pas validées.

Serveur d'exploration sur l'OCR

Adaptive pre-OCR cleanup of grayscale document images

Adaptive pre-OCR cleanup of grayscale document images

Source :

Descripteurs français

English descriptors

Abstract

Links toward previous steps (curation, corpus...)

Le document en format XML

Pour manipuler ce document sous Unix (Dilib)

Pour mettre un lien sur cette page dans le réseau Wicri