Multi-Class Constrained Normalized Cut With Hard, Soft, Unary and Pairwise Priors and its Applications to Object Segmentation

Normalized cut is a powerful method for image segmentation as well as data clustering. However, it does not perform well in challenging segmentation problems, such as segmenting objects in a complex background. Researchers have attempted to incorporate priors or constraints to handle such cases. Ava...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:IEEE transactions on image processing Jg. 22; H. 11; S. 4328 - 4340
Hauptverfasser: Hu, Han, Feng, Jianjiang, Yu, Chuan, Zhou, Jie
Format: Journal Article
Sprache:Englisch
Veröffentlicht: New York, NY IEEE 01.11.2013
Institute of Electrical and Electronics Engineers
Schlagworte:
ISSN:1057-7149, 1941-0042, 1941-0042
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Abstract Normalized cut is a powerful method for image segmentation as well as data clustering. However, it does not perform well in challenging segmentation problems, such as segmenting objects in a complex background. Researchers have attempted to incorporate priors or constraints to handle such cases. Available priors in image segmentation problems may be hard or soft, unary or pairwise, but only hard must-link constraints and two-class settings are well studied. The main difficulties may lie in the following aspects: 1) the nontransitive nature of cannot-link constraints makes it hard to use such constraints in multi-class settings and 2) in multi-class or pairwise settings, the output labels have inconsistent representations with given priors, making soft priors difficult to use. In this paper, we propose novel algorithms, which can handle both hard and soft, both unary and pairwise priors in multi-class settings and provide closed form and efficient solutions. We also apply the proposed algorithms to the problem of object segmentation, producing good results by further introducing a spatial regularity term. Experiments show that the proposed algorithms outperform the state-of-the-art algorithms significantly in clustering accuracy. Other merits of the proposed algorithms are also demonstrated.
AbstractList Normalized cut is a powerful method for image segmentation as well as data clustering. However, it does not perform well in challenging segmentation problems, such as segmenting objects in a complex background. Researchers have attempted to incorporate priors or constraints to handle such cases. Available priors in image segmentation problems may be hard or soft, unary or pairwise, but only hard must-link constraints and two-class settings are well studied. The main difficulties may lie in the following aspects: 1) the nontransitive nature of cannot-link constraints makes it hard to use such constraints in multi-class settings and 2) in multi-class or pairwise settings, the output labels have inconsistent representations with given priors, making soft priors difficult to use. In this paper, we propose novel algorithms, which can handle both hard and soft, both unary and pairwise priors in multi-class settings and provide closed form and efficient solutions. We also apply the proposed algorithms to the problem of object segmentation, producing good results by further introducing a spatial regularity term. Experiments show that the proposed algorithms outperform the state-of-the-art algorithms significantly in clustering accuracy. Other merits of the proposed algorithms are also demonstrated.Normalized cut is a powerful method for image segmentation as well as data clustering. However, it does not perform well in challenging segmentation problems, such as segmenting objects in a complex background. Researchers have attempted to incorporate priors or constraints to handle such cases. Available priors in image segmentation problems may be hard or soft, unary or pairwise, but only hard must-link constraints and two-class settings are well studied. The main difficulties may lie in the following aspects: 1) the nontransitive nature of cannot-link constraints makes it hard to use such constraints in multi-class settings and 2) in multi-class or pairwise settings, the output labels have inconsistent representations with given priors, making soft priors difficult to use. In this paper, we propose novel algorithms, which can handle both hard and soft, both unary and pairwise priors in multi-class settings and provide closed form and efficient solutions. We also apply the proposed algorithms to the problem of object segmentation, producing good results by further introducing a spatial regularity term. Experiments show that the proposed algorithms outperform the state-of-the-art algorithms significantly in clustering accuracy. Other merits of the proposed algorithms are also demonstrated.
Normalized cut is a powerful method for image segmentation as well as data clustering. However, it does not perform well in challenging segmentation problems, such as segmenting objects in a complex background. Researchers have attempted to incorporate priors or constraints to handle such cases. Available priors in image segmentation problems may be hard or soft, unary or pairwise, but only hard must-link constraints and two-class settings are well studied. The main difficulties may lie in the following aspects: 1) the nontransitive nature of cannot-link constraints makes it hard to use such constraints in multi-class settings and 2) in multi-class or pairwise settings, the output labels have inconsistent representations with given priors, making soft priors difficult to use. In this paper, we propose novel algorithms, which can handle both hard and soft, both unary and pairwise priors in multi-class settings and provide closed form and efficient solutions. We also apply the proposed algorithms to the problem of object segmentation, producing good results by further introducing a spatial regularity term. Experiments show that the proposed algorithms outperform the state-of-the-art algorithms significantly in clustering accuracy. Other merits of the proposed algorithms are also demonstrated.
Author Zhou, Jie
Feng, Jianjiang
Hu, Han
Yu, Chuan
Author_xml – sequence: 1
  givenname: Han
  surname: Hu
  fullname: Hu, Han
  email: huh04@mails.thu.edu.cn
  organization: State Key Laboratory on Intelligent Technology and Systems, Tsinghua National Laboratory for Information Science and Technology, Department of Automation, Tsinghua University, Beijing, China
– sequence: 2
  givenname: Jianjiang
  surname: Feng
  fullname: Feng, Jianjiang
  email: jfeng@tsinghua.edu.cn
  organization: State Key Laboratory on Intelligent Technology and Systems, Tsinghua National Laboratory for Information Science and Technology, Department of Automation, Tsinghua University, Beijing, China
– sequence: 3
  givenname: Chuan
  surname: Yu
  fullname: Yu, Chuan
  email: chuanyu08@mails.thu.edu.cn
  organization: State Key Laboratory on Intelligent Technology and Systems, Tsinghua National Laboratory for Information Science and Technology, Department of Automation, Tsinghua University, Beijing, China
– sequence: 4
  givenname: Jie
  surname: Zhou
  fullname: Zhou, Jie
  email: jzhou@tsinghua.edu.cn
  organization: State Key Laboratory on Intelligent Technology and Systems, Tsinghua National Laboratory for Information Science and Technology, Department of Automation, Tsinghua University, Beijing, China
BackLink http://pascal-francis.inist.fr/vibad/index.php?action=getRecordDetail&idt=28088205$$DView record in Pascal Francis
https://www.ncbi.nlm.nih.gov/pubmed/23846473$$D View this record in MEDLINE/PubMed
BookMark eNp9kc1rVDEUxYNU7IfuBUGyEVz0jfl6ycuyPNQWqh1oi8uQSTKakpdMkzxEwf_dTGeq4MJVLje_cy73nmNwEFN0ALzEaIExku9uLpYLgjBdECLwwPsn4AhLhjuEGDloNepFJzCTh-C4lDuEMOsxfwYOCR0YZ4IegV-f5lB9NwZdChxTLDVrH52Fn1OedPA_WznOFX7x9Rs819mewuu0rqfwNur8A-po4VL7_N0XB5fZp1weer4WeLbZBG909c0V1gSvVnfOVHjtvk4u1of-c_B0rUNxL_bvCbj98P5mPO8urz5ejGeXnaFM1s5gKQgbtLHcSWap4VQiLIiW1iI2CLlylknO5PZHG9Ko3iJjmwYNtAlOwNud7yan-9mVqiZfjAtBR5fmojCjTDDKkWjo6z06ryZn1Sb7qW2qHk_WgDd7QBejwzrraHz5y7WRA0F94_iOMzmVkt1aGb_benvioDBS2wxVy1BtM1T7DJsQ_SN89P6P5NVO4p1zf3De9xQTQn8DoZClrw
CODEN IIPRE4
CitedBy_id crossref_primary_10_1109_TIP_2016_2537211
crossref_primary_10_1109_JSTARS_2016_2569408
crossref_primary_10_1109_TPAMI_2014_2377740
crossref_primary_10_3390_rs11010060
crossref_primary_10_1007_s11548_016_1350_2
crossref_primary_10_1007_s13042_015_0458_y
crossref_primary_10_3390_rs10071056
Cites_doi 10.1109/TPAMI.2005.92
10.1109/CVPR.2011.5995630
10.1109/TPAMI.2007.1061
10.1109/CVPR.2005.249
10.1145/321356.321357
10.1145/1015706.1015720
10.1016/j.cviu.2008.06.007
10.1109/MSP.2010.939739
10.1109/TPAMI.2011.231
10.1017/CBO9780511804441
10.1007/s10994-008-5084-4
10.1109/TPAMI.2004.1262179
10.1007/s11222-007-9033-z
10.1111/j.1467-9280.2009.02471.x
10.1007/s11263-009-0275-4
10.1109/34.868688
10.1109/TIP.2009.2038778
10.1109/ICCV.2009.5459173
10.1109/TPAMI.2010.161
10.1109/TPAMI.2007.70840
10.1109/CVPR.2007.382974
10.1109/TPAMI.2009.167
10.1109/ICCV.2007.4408958
10.1145/1835804.1835877
10.1007/s11263-010-0321-2
ContentType Journal Article
Copyright 2015 INIST-CNRS
Copyright_xml – notice: 2015 INIST-CNRS
DBID 97E
RIA
RIE
AAYXX
CITATION
IQODW
CGR
CUY
CVF
ECM
EIF
NPM
7X8
DOI 10.1109/TIP.2013.2271865
DatabaseName IEEE All-Society Periodicals Package (ASPP) 2005–Present
IEEE All-Society Periodicals Package (ASPP) 1998–Present
IEEE/IET Electronic Library
CrossRef
Pascal-Francis
Medline
MEDLINE
MEDLINE (Ovid)
MEDLINE
MEDLINE
PubMed
MEDLINE - Academic
DatabaseTitle CrossRef
MEDLINE
Medline Complete
MEDLINE with Full Text
PubMed
MEDLINE (Ovid)
MEDLINE - Academic
DatabaseTitleList MEDLINE - Academic
MEDLINE

Database_xml – sequence: 1
  dbid: NPM
  name: PubMed
  url: http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=PubMed
  sourceTypes: Index Database
– sequence: 2
  dbid: RIE
  name: IEEE/IET Electronic Library
  url: https://ieeexplore.ieee.org/
  sourceTypes: Publisher
– sequence: 3
  dbid: 7X8
  name: MEDLINE - Academic
  url: https://search.proquest.com/medline
  sourceTypes: Aggregation Database
DeliveryMethod fulltext_linktorsrc
Discipline Applied Sciences
Engineering
EISSN 1941-0042
EndPage 4340
ExternalDocumentID 23846473
28088205
10_1109_TIP_2013_2271865
6553122
Genre orig-research
Research Support, Non-U.S. Gov't
Journal Article
GroupedDBID ---
-~X
.DC
0R~
29I
4.4
53G
5GY
5VS
6IK
97E
AAJGR
AARMG
AASAJ
AAWTH
ABAZT
ABFSI
ABQJQ
ABVLG
ACGFO
ACGFS
ACIWK
AENEX
AETIX
AGQYO
AGSQL
AHBIQ
AI.
AIBXA
AKJIK
AKQYR
ALLEH
ALMA_UNASSIGNED_HOLDINGS
ASUFR
ATWAV
BEFXN
BFFAM
BGNUA
BKEBE
BPEOZ
CS3
DU5
E.L
EBS
EJD
F5P
HZ~
H~9
ICLAB
IFIPE
IFJZH
IPLJI
JAVBF
LAI
M43
MS~
O9-
OCL
P2P
RIA
RIE
RNS
TAE
TN5
VH1
AAYXX
CITATION
AAYOK
IQODW
RIG
CGR
CUY
CVF
ECM
EIF
NPM
7X8
ID FETCH-LOGICAL-c349t-c197248acd6e94d3c6390172a9dd04879bed496493c63ac26e95d0cd2480834d3
IEDL.DBID RIE
ISICitedReferencesCount 9
ISICitedReferencesURI http://www.webofscience.com/api/gateway?GWVersion=2&SrcApp=Summon&SrcAuth=ProQuest&DestLinkType=CitingArticles&DestApp=WOS_CPL&KeyUT=000324597800014&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D
ISSN 1057-7149
1941-0042
IngestDate Mon Sep 29 05:37:10 EDT 2025
Mon Jul 21 06:04:49 EDT 2025
Wed Apr 02 07:25:24 EDT 2025
Tue Nov 18 21:39:51 EST 2025
Sat Nov 29 08:03:37 EST 2025
Wed Aug 27 02:03:32 EDT 2025
IsPeerReviewed true
IsScholarly true
Issue 11
Keywords Performance evaluation
Automatic classification
State of the art
Image processing
Background
pairwise priors
Pattern recognition
Shape detection
Algorithm
Signal classification
Image segmentation
Accuracy
object segmentation
unary priors
Constrained spectral clustering
Edge detection
Language English
License https://ieeexplore.ieee.org/Xplorehelp/downloads/license-information/IEEE.html
https://doi.org/10.15223/policy-029
https://doi.org/10.15223/policy-037
CC BY 4.0
LinkModel DirectLink
MergedId FETCHMERGED-LOGICAL-c349t-c197248acd6e94d3c6390172a9dd04879bed496493c63ac26e95d0cd2480834d3
Notes ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 23
PMID 23846473
PQID 1434743607
PQPubID 23479
PageCount 13
ParticipantIDs pascalfrancis_primary_28088205
pubmed_primary_23846473
crossref_citationtrail_10_1109_TIP_2013_2271865
crossref_primary_10_1109_TIP_2013_2271865
proquest_miscellaneous_1434743607
ieee_primary_6553122
PublicationCentury 2000
PublicationDate 2013-11-01
PublicationDateYYYYMMDD 2013-11-01
PublicationDate_xml – month: 11
  year: 2013
  text: 2013-11-01
  day: 01
PublicationDecade 2010
PublicationPlace New York, NY
PublicationPlace_xml – name: New York, NY
– name: United States
PublicationTitle IEEE transactions on image processing
PublicationTitleAbbrev TIP
PublicationTitleAlternate IEEE Trans Image Process
PublicationYear 2013
Publisher IEEE
Institute of Electrical and Electronics Engineers
Publisher_xml – name: IEEE
– name: Institute of Electrical and Electronics Engineers
References ref35
ref13
ref34
ref12
ref15
ref31
ref30
lu (ref16) 2008
xu (ref18) 2009
ref33
ref11
ref32
ref10
ref2
ref1
ref17
kamvar (ref14) 2003
frank (ref27) 2010
brox (ref4) 2010; 5
li (ref19) 2009
ref24
ref26
zelnik-manor (ref29) 2004
ref25
ref20
ref21
ref28
hu (ref22) 2012
ref8
sundaram (ref7) 2010; 1
ref9
ref3
ref6
cormen (ref23) 2001
ref5
References_xml – year: 2001
  ident: ref23
  publication-title: Introduction to Algorithms
– ident: ref28
  doi: 10.1109/TPAMI.2005.92
– start-page: 421
  year: 2009
  ident: ref19
  article-title: Constrained clustering via spectral regularization
  publication-title: Proc CVPR
– ident: ref21
  doi: 10.1109/CVPR.2011.5995630
– ident: ref24
  doi: 10.1109/TPAMI.2007.1061
– ident: ref3
  doi: 10.1109/CVPR.2005.249
– ident: ref26
  doi: 10.1145/321356.321357
– ident: ref5
  doi: 10.1145/1015706.1015720
– start-page: 561
  year: 2003
  ident: ref14
  article-title: Spectral learning
  publication-title: Proc IJCAI
– ident: ref25
  doi: 10.1016/j.cviu.2008.06.007
– start-page: 1550
  year: 2012
  ident: ref22
  article-title: Multi-way constrained spectral clustering by nonnegative restriction
  publication-title: Proc ICPR
– ident: ref9
  doi: 10.1109/MSP.2010.939739
– volume: 5
  start-page: 282
  year: 2010
  ident: ref4
  article-title: Object segmentation by long term analysis of point trajectories
  publication-title: Proc ECCV
– ident: ref34
  doi: 10.1109/TPAMI.2011.231
– year: 2010
  ident: ref27
  publication-title: UCI Machine Learning Repository
– ident: ref35
  doi: 10.1017/CBO9780511804441
– ident: ref15
  doi: 10.1007/s10994-008-5084-4
– start-page: 1
  year: 2008
  ident: ref16
  article-title: Constrained spectral clustering through affinity propagation
  publication-title: Proc IEEE Conf CVPR
– start-page: 2866
  year: 2009
  ident: ref18
  article-title: Fast normalized cut with linear constraints
  publication-title: Proc IEEE Conf CVPR
– ident: ref2
  doi: 10.1109/TPAMI.2004.1262179
– ident: ref13
  doi: 10.1007/s11222-007-9033-z
– ident: ref8
  doi: 10.1111/j.1467-9280.2009.02471.x
– ident: ref31
  doi: 10.1007/s11263-009-0275-4
– year: 2004
  ident: ref29
  publication-title: Advances in neural information processing systems
– ident: ref11
  doi: 10.1109/34.868688
– ident: ref10
  doi: 10.1109/TIP.2009.2038778
– ident: ref30
  doi: 10.1109/ICCV.2009.5459173
– ident: ref1
  doi: 10.1109/TPAMI.2010.161
– volume: 1
  start-page: 438
  year: 2010
  ident: ref7
  article-title: Dense point trajectories by GPU-accelerated large displacement optical flow
  publication-title: Proc ECCV
– ident: ref33
  doi: 10.1109/TPAMI.2007.70840
– ident: ref32
  doi: 10.1109/CVPR.2007.382974
– ident: ref12
  doi: 10.1109/TPAMI.2009.167
– ident: ref17
  doi: 10.1109/ICCV.2007.4408958
– ident: ref20
  doi: 10.1145/1835804.1835877
– ident: ref6
  doi: 10.1007/s11263-010-0321-2
SSID ssj0014516
Score 2.1649091
Snippet Normalized cut is a powerful method for image segmentation as well as data clustering. However, it does not perform well in challenging segmentation problems,...
SourceID proquest
pubmed
pascalfrancis
crossref
ieee
SourceType Aggregation Database
Index Database
Enrichment Source
Publisher
StartPage 4328
SubjectTerms Accuracy
Algorithms
Applied sciences
Clustering algorithms
Constrained spectral clustering
Correlation
Exact sciences and technology
Image Enhancement - methods
Image Interpretation, Computer-Assisted - methods
Image processing
Image segmentation
Information, signal and communications theory
Matrix decomposition
Motion segmentation
Object detection
Object segmentation
Optimization
pairwise priors
Pattern recognition
Pattern Recognition, Automated - methods
Photography - methods
Reproducibility of Results
Sensitivity and Specificity
Signal and communications theory
Signal processing
Signal representation. Spectral analysis
Signal, noise
Subtraction Technique
Synthetic data
Telecommunications and information theory
unary priors
Title Multi-Class Constrained Normalized Cut With Hard, Soft, Unary and Pairwise Priors and its Applications to Object Segmentation
URI https://ieeexplore.ieee.org/document/6553122
https://www.ncbi.nlm.nih.gov/pubmed/23846473
https://www.proquest.com/docview/1434743607
Volume 22
WOSCitedRecordID wos000324597800014&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D
hasFullText 1
inHoldings 1
isFullTextHit
isPrint
journalDatabaseRights – providerCode: PRVIEE
  databaseName: IEEE/IET Electronic Library
  customDbUrl:
  eissn: 1941-0042
  dateEnd: 99991231
  omitProxy: false
  ssIdentifier: ssj0014516
  issn: 1057-7149
  databaseCode: RIE
  dateStart: 19920101
  isFulltext: true
  titleUrlDefault: https://ieeexplore.ieee.org/
  providerName: IEEE
link http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwlV1RaxQxEB5q8UEfrLZqr-oRwRfhthd395LLYyktCnIetMV7W7JJti7U3bK7pyD0vzuTza0tVMG3sMmwITOZfMlk8gG8c3zGDXexz1iO0Eu6SIu4iKTKFeJTmZv-yfzPcrGYr1ZquQWTIRfGOecvn7lDKvpYvq3Nmo7KpmKGFhOjw30gpehztYaIARHO-sjmTEYSYf8mJMnV9PzTku5wJYdxjJ5YEFkNLlSpSGVyZzXy9Cp0OVK3OD5FT2zxd-TpV6DTnf_r-1N4EpAmO-pN4xlsuWoXdgLqZGFOt7vw-NaThHtw4zNyI8-VyYjN03NIoMCCwO1V-QuLx-uOfS27b4zC_hN2hp58wi4or5fpyrKlLpufZevYsinrpvXfyq5lR7di5ayr2ZeczoDYmbv8HhKgqudwcXpyfvwxChQNkUlS1UWGWMvSuTZWOJXaxAg6Q5GxVtaib0CNO5sqkSqq0SbGVjPLjUUZxH4o8AK2q7py-8AQOUhjrcwlShrxQfPCFLnSQqLqCl2MYLpRVWbC--U0BFeZ38dwlaGeM9JzFvQ8gveDxHX_dsc_2u6RzoZ2QV0jGN-xhqE-ntPehKPc2415ZDgxKdqiK1evW9xTJSnCM8HlCF72dvNHOpjfwf1_fQWPqG99yuNr2O6atXsDD82PrmybMVr_aj721v8b7Rj9-A
linkProvider IEEE
linkToHtml http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwlV3daxQxEB9KFdQHq60f50eN4Itw24u7e8nlsRRLi-d50Cv2bckmWV2ou7K7pyD4vzuTza0tqOBb2GTYkJlMfslk8gN45fiUG-5in7EcoZd0kRZxEUmVK8SnMjf9k_lzuVjMLi7UcgvGQy6Mc85fPnMHVPSxfFubNR2VTcQULSZGh3uDmLNCttYQMyDKWR_bnMpIIvDfBCW5mqxOl3SLKzmIY_TFguhqcKlKRSqTa-uRJ1ih65G6xREqemqLv2NPvwYd7_xf7-_B3YA12WFvHPdhy1W7sBNwJwuzut2FO1ceJdyDnz4nN_JsmYz4PD2LBAosCN5elj-weLTu2Mey-8wo8D9mZ-jLx-ycMnuZrixb6rL5XraOLZuyblr_rexadnglWs66mn3I6RSInblPX0IKVPUAzo_fro5OokDSEJkkVV1kiLcsnWljhVOpTYygUxQZa2UtegfUubOpEqmiGm1ibDW13FiUQfSHAg9hu6or9xgYYgdprJW5REkj3mhemCJXWkhUXaGLEUw2qspMeMGchuAy8zsZrjLUc0Z6zoKeR_B6kPjav97xj7Z7pLOhXVDXCPavWcNQH89od8JR7uXGPDKcmhRv0ZWr1y3uqpIUAZrgcgSPerv5LR3M78mf__oCbp2s3s-z-eni3VO4Tf3sEyCfwXbXrN1zuGm-dWXb7Ps58AtNlABo
openUrl ctx_ver=Z39.88-2004&ctx_enc=info%3Aofi%2Fenc%3AUTF-8&rfr_id=info%3Asid%2Fsummon.serialssolutions.com&rft_val_fmt=info%3Aofi%2Ffmt%3Akev%3Amtx%3Ajournal&rft.genre=article&rft.atitle=Multi-class+constrained+normalized+cut+with+hard%2C+soft%2C+unary+and+pairwise+priors+and+its+applications+to+object+segmentation&rft.jtitle=IEEE+transactions+on+image+processing&rft.au=Hu%2C+Han&rft.au=Feng%2C+Jianjiang&rft.au=Yu%2C+Chuan&rft.au=Zhou%2C+Jie&rft.date=2013-11-01&rft.eissn=1941-0042&rft.volume=22&rft.issue=11&rft.spage=4328&rft_id=info:doi/10.1109%2FTIP.2013.2271865&rft_id=info%3Apmid%2F23846473&rft.externalDocID=23846473
thumbnail_l http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/lc.gif&issn=1057-7149&client=summon
thumbnail_m http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/mc.gif&issn=1057-7149&client=summon
thumbnail_s http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/sc.gif&issn=1057-7149&client=summon