Combinatorial image analysis 19th international workshop iwcia 2018 porto portugal november 22 24 20

Page 1


Combinatorial Image Analysis 19th International Workshop IWCIA 2018

Porto Portugal November 22 24 2018

Proceedings Reneta P. Barneva

Visit to download the full and correct content document: https://textbookfull.com/product/combinatorial-image-analysis-19th-international-work shop-iwcia-2018-porto-portugal-november-22-24-2018-proceedings-reneta-p-barneva /

More products digital (pdf, epub, mobi) instant download maybe you interests ...

Combinatorial Image Analysis 20th International Workshop IWCIA 2020 Novi Sad Serbia July 16 18 2020 Proceedings Tibor Luki■

https://textbookfull.com/product/combinatorial-imageanalysis-20th-international-workshop-iwcia-2020-novi-sad-serbiajuly-16-18-2020-proceedings-tibor-lukic/

Sensor Networks 6th International Conference SENSORNETS 2017 Porto Portugal February 19 21 2017 and 7th

International Conference SENSORNETS 2018 Funchal Madeira Portugal January 22 24 2018 Revised Selected Papers César Benavente-Peces https://textbookfull.com/product/sensor-networks-6thinternational-conference-sensornets-2017-porto-portugalfebruary-19-21-2017-and-7th-international-conferencesensornets-2018-funchal-madeira-portugaljanuary-22-24-2018-revised-sel/

Software Analysis Testing and Evolution 8th International Conference SATE 2018 Shenzhen Guangdong China November 23 24 2018 Proceedings Lei Bu

https://textbookfull.com/product/software-analysis-testing-andevolution-8th-international-conference-sate-2018-shenzhenguangdong-china-november-23-24-2018-proceedings-lei-bu/

Computational Pathology and Ophthalmic Medical Image

Analysis First International Workshop COMPAY 2018 and 5th International Workshop OMIA 2018 Held in Conjunction with MICCAI 2018 Granada Spain September 16 20 2018 Proceedings Danail Stoyanov https://textbookfull.com/product/computational-pathology-andophthalmic-medical-image-analysis-first-international-workshopcompay-2018-and-5th-international-workshop-omia-2018-held-inconjunction-with-miccai-2018-granada-spain-septe/

Knowledge and Systems Sciences 19th International Symposium KSS 2018 Tokyo Japan November 25 27 2018 Proceedings Jian Chen

https://textbookfull.com/product/knowledge-and-systemssciences-19th-international-symposium-kss-2018-tokyo-japannovember-25-27-2018-proceedings-jian-chen/

Image Analysis and Recognition 11th International Conference ICIAR 2014 Vilamoura Portugal October 22 24

2014 Proceedings Part II 1st Edition Aurélio Campilho

https://textbookfull.com/product/image-analysis-andrecognition-11th-international-conference-iciar-2014-vilamouraportugal-october-22-24-2014-proceedings-part-ii-1st-editionaurelio-campilho/

Image Analysis and Recognition 11th International Conference ICIAR 2014 Vilamoura Portugal October 22 24

2014 Proceedings Part I 1st Edition Aurélio Campilho

https://textbookfull.com/product/image-analysis-andrecognition-11th-international-conference-iciar-2014-vilamouraportugal-october-22-24-2014-proceedings-part-i-1st-editionaurelio-campilho/

Product Focused Software Process Improvement 19th International Conference PROFES 2018 Wolfsburg Germany November 28 30 2018 Proceedings Marco Kuhrmann

https://textbookfull.com/product/product-focused-softwareprocess-improvement-19th-international-conferenceprofes-2018-wolfsburg-germany-november-28-30-2018-proceedingsmarco-kuhrmann/

Image Processing and Communications Challenges 10: 10th International Conference, IP&C’2018 Bydgoszcz, Poland, November 2018, Proceedings Micha■ Chora■

https://textbookfull.com/product/image-processing-andcommunications-challenges-10-10th-international-conferenceipc2018-bydgoszcz-poland-november-2018-proceedings-michal-choras/

Valentin E. Brimkov

Combinatorial Image Analysis

19th International Workshop, IWCIA 2018 Porto, Portugal, November 22–24, 2018

Proceedings

LectureNotesinComputerScience11255

CommencedPublicationin1973

FoundingandFormerSeriesEditors: GerhardGoos,JurisHartmanis,andJanvanLeeuwen

EditorialBoard

DavidHutchison

LancasterUniversity,Lancaster,UK

TakeoKanade

CarnegieMellonUniversity,Pittsburgh,PA,USA

JosefKittler UniversityofSurrey,Guildford,UK

JonM.Kleinberg

CornellUniversity,Ithaca,NY,USA

FriedemannMattern

ETHZurich,Zurich,Switzerland

JohnC.Mitchell

StanfordUniversity,Stanford,CA,USA

MoniNaor

WeizmannInstituteofScience,Rehovot,Israel

C.PanduRangan

IndianInstituteofTechnologyMadras,Chennai,India

BernhardSteffen

TUDortmundUniversity,Dortmund,Germany

DemetriTerzopoulos UniversityofCalifornia,LosAngeles,CA,USA

DougTygar UniversityofCalifornia,Berkeley,CA,USA

GerhardWeikum

MaxPlanckInstituteforInformatics,Saarbrücken,Germany

Moreinformationaboutthisseriesat http://www.springer.com/series/7411

Combinatorial ImageAnalysis

19thInternationalWorkshop,IWCIA2018

Porto,Portugal,November22–24,2018

Proceedings

Editors

RenetaP.Barneva StateUniversityofNewYorkatFredonia Fredonia,NY,USA

ValentinE.Brimkov StateUniversityofNewYork Buffalo,NY,USA

and InstituteofMathematicsandInformatics

BulgarianAcademyofSciences Sofia,Bulgaria

ISSN0302-9743ISSN1611-3349(electronic) LectureNotesinComputerScience

ISBN978-3-030-05287-4ISBN978-3-030-05288-1(eBook) https://doi.org/10.1007/978-3-030-05288-1

LibraryofCongressControlNumber:2018962921

LNCSSublibrary:SL5 – ComputerCommunicationNetworksandTelecommunications

© SpringerNatureSwitzerlandAG2018

Thisworkissubjecttocopyright.AllrightsarereservedbythePublisher,whetherthewholeorpartofthe materialisconcerned,specificallytherightsoftranslation,reprinting,reuseofillustrations,recitation, broadcasting,reproductiononmicrofilmsorinanyotherphysicalway,andtransmissionorinformation storageandretrieval,electronicadaptation,computersoftware,orbysimilarordissimilarmethodologynow knownorhereafterdeveloped.

Theuseofgeneraldescriptivenames,registerednames,trademarks,servicemarks,etc.inthispublication doesnotimply,evenintheabsenceofaspecificstatement,thatsuchnamesareexemptfromtherelevant protectivelawsandregulationsandthereforefreeforgeneraluse.

Thepublisher,theauthors,andtheeditorsaresafetoassumethattheadviceandinformationinthisbookare believedtobetrueandaccurateatthedateofpublication.Neitherthepublishernortheauthorsortheeditors giveawarranty,expressorimplied,withrespecttothematerialcontainedhereinorforanyerrorsor omissionsthatmayhavebeenmade.Thepublisherremainsneutralwithregardtojurisdictionalclaimsin publishedmapsandinstitutionalaffiliations.

ThisSpringerimprintispublishedbytheregisteredcompanySpringerNatureSwitzerlandAG Theregisteredcompanyaddressis:Gewerbestrasse11,6330Cham,Switzerland

Preface

Thisvolumecontainstheproceedingsofthe19thInternationalWorkshoponCombinatorialImageAnalysis(IWCIA2018)heldinPorto,Portugal,November22–24, 2018.

Imageanalysisprovidestheoreticalfoundationsandmethodsforsolvingreal-life problemsarisinginvariousareasofhumanpractice,suchasmedicine,robotics, defense,andsecurity.Sincetypicallytheinputdatatobeprocessedarediscrete,the “combinatorial ” approachtoimageanalysisisanaturaloneandthereforeitsapplicabilityisexpanding.Combinatorialimageanalysisoftenprovidesadvantagesinterms ofefficiencyandaccuracyoverthemoretraditionalapproachesbasedoncontinuous modelsthatrequirenumericcomputation.

TheIWCIAworkshopseriesprovidesaforumforresearchersthroughouttheworld topresentcutting-edgeresultsincombinatorialimageanalysis,todiscussrecent advancesandnewchallengesinthisresearcharea,andtopromoteinteractionwith researchersfromothercountries.IWCIAhadsuccessfulpriormeetingsinParis (France)1991,Ube(Japan)1992,WashingtonDC(USA)1994,Lyon(France)1995, Hiroshima(Japan)1997,Madras(India)1999,Caen(France)2000,Philadelphia,PA (USA)2001,Palermo(Italy)2003,Auckland(NewZealand)2004,Berlin(Germany) 2006,Buffalo,NY(USA)2008,PlayadelCarmen(Mexico)2009,Madrid(Spain) 2011,Austin,TX(USA)2012,Brno(CzechRepublic)2014,Kolkata(India)2015,and Plovdiv(Bulgaria)2017.TheworkshopinPortoretainedandenrichedtheinternational spiritoftheseworkshops.TheIWCIA2018ProgramCommitteemembersare renownedexpertscomingfrom17differentcountriesinAsia,Europe,NorthandSouth America,andtheauthorscomefrom12differentcountries.

Eachsubmittedpaperwassenttoatleastthreereviewersforadouble-blindreview. EasyChairprovidedaconvenientplatformforsmoothlycarryingouttherigorous reviewprocess.Themostimportantselectioncriterionforacceptanceorrejectionofa paperwastheoverallscorereceived.Othercriteriaincluded:relevancetotheworkshop topics,correctness,originality,mathematicaldepth,clarity,andpresentationquality. Webelievethatasaresultonlypapersofhighqualitywereacceptedforpresentationat IWCIA2018andforpublicationinthisvolume.

Theprogramoftheworkshopincludedpresentationsofcontributedpapersand keynotetalksbythreedistinguishedscientists.ThetalkofJosé M.Bioucas-Dias (UniversityofLisbon,Portugal)presentedmethodsforsolvingcertaininverseproblemsofinterferometricphaseimaging,aimedattheestimationofphasefromsinusoidal andnoisyobservations.Heproposedcharacterizationsoftheobtainedestimatesand illustratedtheireffectivenessbyresultsofextensiveexperimentswithsimulatedand realdata.JanKratochvil(CharlesUniversity,CzechRepublic)reviewedvarious problemsonintersection,contact,orvisibilityrepresentationsofgraphs,whicharea naturalwayofgraphandnetworkvisualization.Hesurveyedknownresultsonthese classesofgraphsanddiscussednewaswellaspersistingopenproblems.Nicolai

Petkov(UniversityofGroningen,TheNetherlands)presentednoveltrainable filtersas analternativetodeepnetworksfortheextractionofeffectiverepresentationsofdatafor thepurposesofpatternrecognition.Hediscussedtheadvantagesofthisapproachand illustrateditonhandwrittendigitimages,trafficsigns,andsoundsinaudiosignals.

Thecontributedpapersaregroupedintotwosections.The firstoneincludesnine papersdevotedtotheoreticalfoundationsofcombinatorialimageanalysis,including digitalgeometryandtopology,arraygrammars,tilingsandpatterns,discretegeometry innon-rectangulargrids,andothertechnicaltoolsforimageanalysis.Thesecondpart includesninepaperspresentingapplication-drivenresearchontopicssuchasdiscrete tomography,imagesegmentation,textureanalysis,andmedicalimaging.Webelieve thatmanyofthesepaperswouldbeofinteresttoabroaderaudience,including researchersinscientificareassuchascomputervision,shapemodeling,patternanalysis andrecognition,andcomputergraphics.

Aspecialsessionprovidedsomeauthorswiththeopportunitytopresenttheir ongoingresearchprojectsandoriginalworksinprogress.Thetextsoftheseworksare notincludedinthisvolume.

ManyindividualsandorganizationscontributedtothesuccessofIWCIA2018.The editorsareindebtedtoIWCIA’sSteeringCommitteeforendorsingthecandidacyof Portoforthe19theditionoftheworkshop.Wewishtothankeverybodywhosubmitted theirworktoIWCIA2018;thankstotheircontributions,wesucceededinhavinga technicalprogramofhighscientifi cquality.Wearegratefultoallparticipantsand especiallytothecontributorsofthisvolume.OurmostsincerethanksgototheIWCIA 2018ProgramCommitteewhosecooperationincarryingouthigh-qualityreviewswas essentialinestablishingastrongscienti ficprogram.Weexpressoursinceregratitudeto thekeynotespeakers,José M.Bioucas-Dias,JanKratochvil,andNicolaiPetkovfor theirexcellenttalksandoverallcontributiontotheworkshopprogram.

Thesuccessoftheworkshopwouldnotbepossiblewithoutthehardworkofthe localOrganizingCommittee.Wearegratefultoourpartners:FEUP – Faculdadede EngenhariadaUniversidadedoPorto;LABIOMEP:LaboratóriodeBiomecânicado Porto;APMTAC – AssociaçãoPortuguesadeMecânicaTeórica,AplicadaeComputacional;FCT – FundaçãoparaaCiênciaeaTecnologia;INEGI – Institutode CiênciaeInovaçãoemEngenhariaMecânicaeEngenhariaIndustrial;ITHEAISS –InternationalScientificSociety;andTaylor&FrancisGroup.Finally,wewishtothank Springer,andespeciallyAlfredHofmannandAnnaKramer,fortheireffi cientandkind cooperationinthetimelyproductionofthisbook.

October2018RenetaP.Barneva ValentinE.Brimkov JoãoManuelR.S.Tavares

Organization

The19thInternationalWorkshoponCombinatorialImageAnalysis,IWCIA2019was heldinPorto,Portugal,November22–24,2018.

GeneralChair

JoãoManuelR.S.TavaresUniversityofPorto,Portugal Chairs

RenetaP.BarnevaSUNYFredonia,USA ValentinE.BrimkovSUNYBuffaloState,USA RenatoM.NatalJorgeUniversityofPorto,Portugal

SteeringCommittee

BhargabB.BhattacharyaIndianStatisticalInstitute,Kolkata,India ValentinE.BrimkovSUNYBuffaloState,Buffalo,USA GaborHermanCUNYGraduateCenter,USA KostadinKorutchevUniversidadAutonoma,Spain JosefSlapalTechnicalUniversityofBrno,CzechRepublic

InvitedSpeakers

José M.Bioucas-DiasUniversityofLisbon,Portugal JanKratochvilCharlesUniversity,CzechRepublic NicolaiPetkovUniversityofGroningen,TheNetherlands ProgramCommittee

EricAndresUniversité dePoitiers,France PeterBalazsUniversityofSzeged,Hungary JackyBaltesUniversityofManitoba,Canada GeorgeBebisUniversityofNevadaatReno,USA BhargabB.BhattacharyaIndianStatisticalInstitute,India ParthaBhowmickIITKharagpur,India ArindamBiswasIIESTShibpur,India PeterBrassCityCollege,CityUniversityofNewYork,USA BorisBrimkovRiceUniversity,USA SreckoBrlekUniversité duQuebecaMontreal,Canada AlfredM.BrucksteinTechnion,I.I.T,Israel

Jean-MarcChasseryUniversité deGrenoble,France

LiChenUniversityoftheDistrictofColumbia,USA

DavidCoeurjollyUniversité deLyon,France

RocioGonzalez-DiazUniversityofSeville,Spain MousumiDuttSt.ThomasCollegeofEngineeringandTechnology, India

FabienFeschetUniversité d’Auvergne,France

LeilaDeFlorianiUniversityofGenova,ItalyandUniversityof Maryland,USA

Chiou-ShannFuhNationalTaiwanUniversity,Taiwan EdwinHancockUniversityofYork,UK

FayHuangNationalIlanUniversity,Taiwan KrassimiraIvanovaBulgarianAcademyofSciences,Bulgaria

AtsushiImiyaIMIT,ChibaUniversity,Japan

MariaJoseJimenezUniversityofSeville,Spain

KamenKanevShizuokaUniversity,Japan YungKongCUNYQueensCollege,USA VladimirKovalevskiTechnischeFachhochschuleBerlin,Germany

WalterG.KropatschViennaUniversityofTechnology,Austria

LonginJanLateckiTempleUniversity,USA

JeromeLiangSUNYStonyBrook,USA

JoakimLindbladUppsalaUniversity,Sweden

BenedekNagyEasternMediterraneanUniversity,Cyprus

GregoryM.NielsonArizonaStateUniversity,USA JanosPachEPFLLausanne,SwitzerlandandRenyiInstitute Budapest,Hungary

KalmanPalagyiUniversityofSzeged,Hungary PetraPernerInstituteofComputerVisionandAppliedComputer Sciences,Germany

NicolaiPetkovUniversityofGroningen,TheNetherlands

HemersonPistoriDomBoscoCatholicUniversity,Brazil

IoannisPitasUniversityofThessaloniki,Greece KonradPolthierFreieUniversitätBerlin,Germany HongQinSUNYStonyBrook,USA PaoloRemagninoKingstonUniversity,UK

RalfReulkeHumboldtUniversity,Germany BodoRosenhahnMPIInformatik,Germany ArunRossMichiganStateUniversity,USA NikolaySirakovTexasA&MUniversity-Commerce,USA

RaniSiromoneyMadrasChristianCollege,India WladyslawSkarbekWarsawUniversityofTechnology,Poland AliShokoufandehDrexelUniversity,USA

K.G.SubramanianMadrasChristianCollege,India

GnanarajThomasMadrasChristianCollege,India PeterVeelaertUniversityofGent,Belgium

PetraWiederholdCINVESTAV-IPN,Mexico JinhuiXuSUNYUniversityatBuffalo,USA

LocalOrganizingCommittee

JessicaDelmoralUniversityofPorto,Portugal

MarcoMarquesUniversityofPorto,Portugal

JoãoNunesUniversityofPorto,Portugal

André LuizPilastriUniversityofPorto,Portugal DomingosVieiraUniversityofPorto,Portugal

AdditionalReviewers

ChouaibMoujahdi

TinNguyen AliErol JiriHladuvka MinweiYe XiangyuWang UldericoFugacci PaolaMagillo ZiyunHuang FedericoIuricich

Contents

TheoreticalContributions

OnGapsinDigitalObjects...................................3

Lidija Čomić

FixpointsofIteratedReductionswithEquivalentDeletionRules..........17 KálmánPalágyiandGáborNémeth

ParallelContextualArrayInsertionDeletionGrammar.................28

D.G.Thomas,S.JamesImmanuel,AtulyaK.Nagar, andRobinsonThamburaj

ChainCodePSystemforGenerationofApproximationPatterns ofSierpi ńskiCurve.........................................43

A.Dharani,R.StellaMaragatham,AtulyaK.Nagar, andK.G.Subramanian

DigitizedRotationsofClosestNeighborhoodontheTriangularGrid.......53 AydınAvkan,BenedekNagy,andMügeSaadetoğlu

BinaryTomographyonTriangularGridInvolvingHexagonal GridApproach............................................68

BenedekNagyandTiborLukić

SphereConstructionontheFCCGridInterpretedasLayered HexagonalGridsin3D......................................82

GirishKoshti,RanitaBiswas,GaëlleLargeteau-Skapin,RitaZrour, EricAndres,andParthaBhowmick

QuadrangularMeshGenerationUsingCentroidalVoronoiTessellation onVoxelizedSurface.......................................97 AshutoshSoniandParthaBhowmick

WhenCan lp -normObjectiveFunctionsBeMinimizedviaGraphCuts?....112 FilipMalmbergandRobinStrand

FromTheorytoPractice

DetectionofOsteoarthritisbyGapandShapeAnalysisofKnee-Bone X-ray...121 SabyasachiMukherjee,OishilaBandyopadhyay,ArindamBiswas, andBhargabB.Bhattacharya

RegistrationofCTwithPET:AComparisonofIntensity-Based Approaches..............................................134

GisèlePereira,InêsDomingues,PedroMartins,PedroH.Abreu, HugoDuarte,andJoãoSantos

IterativeHighResolutionTomographyfromCombinedHigh-Low ResolutionSinogramPairs....................................150

László VargaandRajmundMokso

AMulti-channelDARTAlgorithm..............................164

Mathé Zeegers,FelixLucka,andKeesJoostBatenburg

EvaluationofChaosGameRepresentationforComparison ofDNASequences.........................................179

André R.S.Marcal

Image-BasedObjectSpoofingDetection..........................189

ValterCosta,ArmandoSousa,andAnaReis

ImprovementofMeasurementAccuracyofOptical3DScanners byDiscreteSystematicErrorEstimation...........................202

ChristianBräuer-Burchardt,PeterKühmstedt,andGuntherNotni

DefectDetectioninTextileswithCo-occurrenceMatrixasaTexture ModelDescription..........................................216

KarolinaNurzynskaandMichał Czardybon

MultiscaleGraph-Cutfor3DSegmentationofCompactObjects..........227

MiroslavJirik,VladimirLukes,MilosZelezny,andVaclavLiska

AuthorIndex

TheoreticalContributions

OnGapsinDigitalObjects

FacultyofTechnicalSciences,UniversityofNoviSad,NoviSad,Serbia

Abstract. Differentformulaewereproposedintheliteratureforthe numberofgapsindigitalobjects.Wegiveseveralnewformulaeforthe numberof0-gapsin2D,basedontheknownconnectionbetweenthe numberof0-gapsandtheEulercharacteristicof2Ddigitalobjects.We alsopresenttwonew,shortandintuitiveproofsofoneofthetwoknown equivalentformulaeforthenumberof(n 2)-gapsin nDdigitalobjects.

Keywords: Digitalobjects · Digitaltopology · Gaps Eulercharacteristic

1Introduction

Topologicalanalysisofimagesandshapesisanactiveresearchfield,important inmanyapplicationdomains.Twonumericaltopologicaldescriptorsofdigital objects,thenumberofgapsandtheEulercharacteristic,havebeenwidelyinvestigatedintheliterature.Establishingtherelationsbetweenthesedescriptorsand thenumberofcellsofdifferenttypesinthedecompositionofthedigitalobject O intoacellcomplex Q enablesabetterunderstandingandadeeperinsight intothetopologicalstructureoftheobject O andcomplex Q.

Intuitively,gapsinadigitalobject O arelocationsin O throughwhicha discretecurvecanpenetrate[6].Theyplayanimportantroleinray-casting basedrenderingofdigitalcurvesandsurfaces[15, 46],andinparticularofdigitalplanes[1, 24, 30].Thenotionofagapiscloselyrelatedtothenotionof well-composedness[4, 35, 36].Thenumberofgapscharacterizesthetopological structureoftheobject[10].TheEulercharacteristicisanotherbasic,locally computabletopologicaldescriptor.In2D,itisequaltothenumberofconnected componentsoftheobject O minusthenumberofholesin O .

Severaldifferentformulaeforthenumberof0-gapsina2D[9, 10]and1gapsina3Ddigitalobject O [38]havebeenproposed,intermsofthenumber ofcellsinthecubicalcomplex Q associatedwiththeobject O (andtheEuler characteristicof O [9]).

Forthenumberof(n 2)-gapsinan nDdigitalobject O ,therearetwo suchformulae.One[6, 11],expressesthenumberofgapsin O throughthetotal numberofcellsofdimension n, n 1and n 2,andthenumberofinteriorcells ofdimension n 2inthecubicalcomplex Q associatedwith O .Thisformulawas usedtoobtainsomecombinatorialrelationsfordigitalcurves[6, 11].Relations c SpringerNatureSwitzerlandAG2018

R.P.Barnevaetal.(Eds.):IWCIA2018,LNCS11255,pp.3–16,2018. https://doi.org/10.1007/978-3-030-05288-1 1

betweenthedimensionofahypersurface(andinparticularadigitalhyperplane) andthetypesofgapsthehypersurfacemaypossess[6, 8],andgoodpairsof adjacencyrelationsin nD[8]wereestablished.

Theotherformula[39]expressesthenumberof(n 2)-gapsin O through thenumberofboundarycellsin Q ofdimension n 1and n 2.Itisobtained bydefiningseveralnewnotionsinthediscretespaceandderivingmanyinterestingrelationsbetweenvarioustypesofcellsin Q.Thenumberof0-gapsin 2D,obtainedthroughthe2Dspecializationofthisformula,togetherwiththe numberof2-,1-and0-cellsin Q,wasrelatedtothedimensionofthe2Ddigital object O [40].ArelationbetweenthedimensionandtheEulercharacteristicof 2Ddigitalobjectswasalsoestablished[12].

Thecontributionsofthispaperare:

–Severalnewformulaeforthenumberof0-gapsin2D,obtainedfromtheknown formulaefortheEulercharacteristicofa2Ddigitalobject O [25, 43, 45]and therelation[9]betweenthenumberofgapsin O anditsEulercharacteristic.

–Twoalternative,shortandintuitiveproofsforone[39]ofthetwoequivalent formulaeforthenumberof(n 2)-gapsin nDdigitalobjects.Oneproofis combinatorial,andtheotherisgraphtheoretical.

2Preliminaries

Weintroducesomebasicnotionsonthecubicalgrid[31, 32],ongapsindigital objectsinthecubicalgrid[6, 9, 24, 38]andontheEulercharacteristicofsuch objectsin2D[32].

2.1TheCubicalGrid

Definition1. An nDcubicalgridisasetofclosedunit n-cubes(voxels)centeredatpointsin Zn ,withfacesparalleltothecoordinateplanes.Thenaturally associated(cubical)cellcomplexiscomposedofallvoxels,andalltheir k -faces (k -cells), 0 ≤ k ≤ n 1

Differenttypesofadjacencyrelationaredefinedbetweenthevoxels(n-cells)in thegrid,dependingontheirintersection.

Definition2. Twovoxelsare k -adjacentiftheyshareacommon k -cell.They arestrictly k -adjacentiftheyare k -adjacentbutnot (k +1)-adjacent. In2D,twopixels(2-cells)thatshareanedgeoravertexare1-or0-adjacent, respectively.In3D,twovoxelsthatshareasquareface,anedgeoravertexare 2-,1-or0-adjacent,respectively.

Combinatorialcoordinatesystemisnaturallydefinedforthecubicalgrid[33]. Itaddressesgridcellsofalldimensionsandenableseasynavigationbetweenthem throughintegerarithmetics.Wegivesomebasicpropertiesofcellcoordinates andexamplesofthetopologicalrelationsthatwillbeusedinthesequel.To obtainintegercoordinatevaluesforallcells,thegridisrescaledbyfactor2in allcoordinatedirections.

Definition3. EachcellinthecubicalgridisassignedthecombinatorialcoordinatesequaltothedoubleoftheCartesiancoordinatesofits(bary)center.

A k -cellhas k evenand n k oddcoordinates.Avoxel(n-cell)has n even,and avertex(0-cell)has n oddcoordinates.

Thedimension,spatialpositionandrelativeorientationofacellareencoded initscoordinates.Forexample,an n-tupleconsistingofoneevenand n 1 oddnumbersrepresentsanedgeinthecombinatorialcoordinatesystem.The Cartesiancoordinatesofitscenterareobtainedbydividingthe n-tupleby2.If the ithcoordinateiseven,theedgeisparalleltothe ithcoordinateaxis.

Topologicalincidenceandadjacencyrelationsbetweencellscanbeeasily retrievedfromthecoordinatevalues.

Proposition1. Each k -cell e isboundedby 2k i k i i-cells, 0 ≤ i ≤ k ,and bounds 2j k n k j k j -cells, k ≤ j ≤ n.

Theboundarycellsareobtainedbychanging k i evencoordinatesof e by ±1. Thecoboundarycellsareobtainedbychanging j k oddcoordinatesof e by ±1.Forexample,avoxelisboundedby2n (n 1)-cells,thatareobtainedby changingoneofthevoxelcoordinatesby ±1,andfixingtheremaining n 1 coordinates.

An(n 1)-cellisboundedby2(n 1)(n 2)-cells,thatareobtainedbyfixing theoddcoordinateandchangingoneoftheremaining n 1evencoordinates by ±1.Itboundstwovoxels,thatareobtainedbychangingtheoddcoordinate ofthe(n 1)-cellby ±1.

An(n 2)-cellhastwooddcoordinatesand n 2evenones.Itisincident tofourvoxels,thatareobtainedbychangingeachofthetwooddcoordinates by ±1,andtofour(n 1)-cells,thatareobtainedbychangingby ±1oneand keepingtheotherofthetwooddcoordinates.

2.2GapsinDigitalObjects

Definition4. An nD digitalobject O isafinitesetofvoxelsinthecubical grid.Theassociatedcubicalcomplexisdenotedas Q.Itconsistsofallvoxelsin O andalltheir k -dimensionalfaces.

Thenumberof k -faces(k -cells)in Q isdenotedas ck .Thevoxelsin O arecalled black(object)voxels.Thevoxelsinthecomplement O c of O arecalledwhite (background).

A2n k × 1k blockofvoxelsiscalleda2n k 1k -block,0 ≤ k ≤ n.The2n k 1kblock Bk (e)centeredata k -cell e consistsof2n k voxelsincidentto e

Definition5. Apairofstrictly k -adjacentvoxelsthrougha k -cell e iscalleda k -tandem over e, 0 ≤ k ≤ n 1.

Definition6. An nDdigitalobject O hasa k -gap ata k -cell e if Bk (e)\O isa k -tandemover e

Anobject O hasan(n 2)-gapatan(n 2)-cell e iftwo(outoffour)voxels incidentto e (inthe22 1n 2 -block Bn 2 (e))areblackandstrictly(n 2)-adjacent through e (andtheothertwoarewhiteandstrictly(n 2)-adjacentthrough e). Thus,theblackandwhitevoxelsalternatecyclicallyaroundthe(n 2)-gap e.

In2D,a0-gapoccursatavertexthatisincidenttotwostrictly0-adjacent whitepixels(andtwoblackones).In3D,a1-gapoccursatanedgeincidentto twostrictly1-adjacentwhitevoxels(andtwoblackones).A0-gapoccursata vertexincidenttotwostrictly0-adjacentwhitevoxels(andtosixblackones). Thenumberof k -gapsinanobjectisdenotedas gk

Definition7. Aboundary(free) k -cellin Q, 0 ≤ k ≤ n 1,isa k -cellincident bothtoavoxelin O andavoxelin O c .Aninteriorcellisincidenttovoxelsin O only.

A k -cell e in Q isinteriorifthe2n k 1k -block Bk (e)centeredat e iscontained in O .Otherwise, e isaboundary k -cell[10, 39].Thenumberofinteriorand boundary k -cellsin Q isdenotedas ck and c∗ k ,respectively,and

Ifan nDobject O hasan(n 2)-gapatan(n 2)-cell e,then e isaboundary (n 2)-cellin Q,andallfour(n 1)-cellsincidentto e areboundary(n 1)cellsin Q.Foranon-gapboundary(n 2)-cell e in Q,twooutoffourincident (n 1)-cellsareboundary(n 1)-cellsin Q.Theseclaimscanbeverifiedby examiningthepossibleconfigurationsofthefourvoxelsintheblock Bn 2 (e) [11, 39],orbyinspectingthecombinatorialcoordinatesoftheinvolvedcells(see Sect. 2.1).

2.3TheEulerCharacteristicof2DDigitalObjects

TheEulercharacteristicisoneofthebasictopologicaldescriptorsofashape.It canbedefinedintwoequivalentways:throughthetopologicalpropertiesrelated totheconnectednessof O andthroughthenumberofcellsinthecubicalcomplex Q associatedwith O .Ineithercase,itsvaluedependsonthechosenadjacency relation.Wedefineitherefor2Ddigitalobjects.

Twopixels(2-cells) p and q ina2Dobject O are k -connectedin O , k =0, 1, ifthereisasequenceofpixelsin O ,startingat p andendingat q ,suchthatany twoconsecutivepixelsinthesequenceare k -adjacent.Connected k -components of O aremaximalsubsetsof O withrespectto k -connectedness.Theirnumber isdenotedas ck .Aholeisafiniteconnectedcomponentofthecomplement O c of O .Ifconnectedcomponentsof O arecountedwithrespectto0-adjacency, thenholesarecountedwithrespectto1-adjacencyandviceversa.Thenumber ofholeswith k -adjacencyfor O and(1 k )-adjacencyfor O c isdenotedas h1 k . Wedenoteas χk (O )theEulercharacteristicwith k -adjacencyforblackpixels and(1 k )-adjacencyforwhitepixels, k =0, 1.

Definition8. TheEulercharacteristic χk (O ), k =0, 1,ofa2Ddigitalobject O isequaltothenumberofconnectedcomponentsof O minusthenumberof holesin O ,i.e., χ k (O )= c k h1 k .

Fora2Ddigitalobjectwith0-adjacency,i.e.,fortheassociatedcubicalcomplex Q,theEulercharacteristiccanbecomputedasthealternatingsumofthenumber ci of i-cellsin Q, i =0, 1, 2.For1-adjacency,verticesexhibitinga0-gapare duplicated.Thus,

ThecomputationoftheEulercharacteristicofa2Dobject O isawidely investigatedresearchfield.Manyalgorithms,andthecorrespondingformulae, wereproposedintheliterature,bothfor0-andfor1-adjacencyrelation.

SomealgorithmsarebasedonrelationsbetweentheEulercharacteristicand thenumberofcellsofvarioustypesinthecomplex Q [5, 42, 43, 45]orinthedual complex[14, 28, 48].Somealgorithmsexaminetheboundarycurve C (composed ofedgesandvertices)of Q,andcountthenumberofconvexandconcavevertices in C [13, 25, 29, 43, 44].Othersarebasedoncountingdifferentpixelconfigurations aroundtheverticesin Q [25, 47],oronthedivideandconquerstrategy[20, 21]. Someareadaptedtoaspecificdatastructureencodingthecomplex Q,such asquad-tree[23],skeleton[22]orrun-lengthencoding[2, 3, 20, 21, 37, 41, 48, 49]. SomealgorithmsadoptthetopologicaldefinitionoftheEulercharacteristic, anduseaconnectedcomponentlabelingalgorithmforcountingthenumberof connectedcomponentsandthenumberofholes[26, 27].

3RelatedWork

Wereviewtheformulaeproposedintheliteratureforthenumberof0-gapsin 2D,1-gapsin3Dand(n 2)-gapsin nDdigitalobjects.

3.10-Gapsin2DObjects

Twoequivalent[40]formulaewereproposedforthenumber g0 of0-gapsina2D digitalobject O . One[9]statesthat

where c0 isthenumberof0-components(maximalconnectedcomponentsof blackpixelswithrespectto0-adjacency),and h1 isthenumberof1-holes(maximalfiniteconnectedcomponentsofwhitepixelswithrespectto1-adjacency). Theproofisbyinductiononthenumberofpixels.Forthebasecase,the gap-freeobject O consistsofasingleblackpixel,and c0 =4, c2 =1, c0 =1, h1 =0, c0 =0.Fortheinductivestep,allpossibleconfigurationsofblackand

whitepixelsareexaminedina3 × 3neighborhoodofapixeladdedtoadigital objectwith n pixelsforwhichtheformulaisvalid.

Analternativeformulaforthenumberofgapsin2D[10]expresses g0 in termsofboundarycellsin Q as

Theproofisbyinductiononthenumberofpixels,andisbasedonexamining thepossible3 × 3configurationsofpixels.

Anotherproof[7, 12]oftheformula

usesthefactthatthegraphcomposedofboundaryverticesandedgesof O is Eulerian,andthateachgapvertexisincidenttofour,andeachnon-gapvertex isincidenttotwoedges.SinceedgesandverticesalternateontheEuleriancycle,

UsingtheEulerformula

itfollowsthat

Thisrelationwasobtainedindependentlyinthecontextofcalculatingthenumberofholesina2Dbinaryimage[42].Thenumberof0-gapsin2Dwasrelated tothedimensionofadigitalobject O [40].

3.21-Gapsin3DObjects

Oneformulawasproposed[38]forthenumber g1 of1-gapsina3Ddigitalobject O ,namely

Theproofisbasedonconstructingagraph G =(N,A),andusingthatthe sumofvertexdegreesisequaltotwicethenumberofarcs.Thenodesin N correspondtotheboundary2-cellsin Q,andtwonodesin N areconnected throughanarcin A ifthecorresponding2-cells(squarefaces)shareacommon 1-cell(edge).Eachnon-gapboundary1-cellisincidenttotwoboundary2-cells, andcontributes 2 2 =1arcin A.Each1-gapisincidenttofourboundary2cells,andcontributes 4 2 =6arcs.Thedegreeofanodedependsonwhether thecorresponding2-cellisincidenttoa1-gapornot.

3.3(n 2)-Gapsin nDObjects

Twoformulaewereproposedintheliteratureforthenumber gn 2 of(n 2)-gaps inan nDdigitalobject O

One[6, 11]expressesthenumberofgapsthroughthetotalnumberof n-, (n 1)-and(n 2)-cells,andthenumberofinterior(n 2)-cellsin Q as

Theother[39]expressesthisnumberintermsofboundary(n 1)-and(n 2)cellsin Q as

Theproofsofbothformulaerelyontheregularstructureofthecubicalgrid,in whichthenumberofcellsintheboundaryandcoboundaryofagiven k -cellis constant:each n-cellin Q isincidentto2n (n 1)-cellsin Q;each(n 1)-cellis incidenttooneortwo n-cells,dependingifitisaboundaryorinterior(n 1)-cell in Q;each(n 2)-cellin Q isincidentto0(ifitisaninteriorcell),2(ifitisa non-gapcell)or4(ifitisan(n 2)-gap)boundary(n 1)-cells.

Theequivalenceofthetwoformulaewasestablished[39],basedonthefollowingrelation:

Therelationfollowsfromthefactthateachofthe cn n-cells(voxels)in Q is boundedby2n (n 1)-cells.Eachboundary(n 1)-cellisincidenttoonevoxel, andeachinterior(n 1)-cellisincidenttotwovoxels[6, 11](seealsoSect. 2.1). Thus,boundary(n 1)-cellsarecountedonce,andinterior(n 1)-cellsare countedtwotimesin2ncn ,i.e.,

orequivalently

4SomeNewFormulaefortheNumberof0-Gapsin2D

Weproposeseveralnewformulaeforthenumberof0-gapsin2Ddigitalobjects. (1)Weobtainthefirstonefromtheknownformula[43]

fortheEulercharacteristic χ1 ,proposedasanextensiontoarbitraryobjects oftheformula[5]validforunit-width(withoutinnervertices)objectswith 1-adjacency.

Letusconsideramultigraph G =(N,A),whosenodescorrespondtomaximalconnected1-componentsof O ,andthearcscorrespondtogapsatwhich

suchcomponentsare(0-)connected.Letusdenoteas χ1 (Oj )theEulercharacteristicofsuchcomponent Oj .Ifweiterativelyaddthearcsin A tothe nodesin N ,theneacharceitherconnectstwodifferentconnectedcomponentsofnodesin N ,oritcreatesaloopconnectingtwonodesfromthe sameconnectedcomponentofnodesin N (itcreatesaholein G).Ineither case,eacharc(each0-gap)decreasestheEulercharacteristicby1.Thefinal graph G hasthesameEulercharacteristicastheobject O with0-adjacency, i.e.,

(G)= j

and

WeusetheEulercharacteristicformula χ1 = 1 4 (

∗ 1 2c

)+ c0 , andapplyit oneachconnectedcomponent Oj .Usingthefactthateachinnerorboundary cellin O isinnerorboundarycellinsomecomponent Oj ,wehave g0 = j ( 1 4 (

Usingthedefinitionof χ0 asthealternatingsumof i-cells, i =0, 1, 2,we get

Thesameformulafollowsfrom

[7, 12]and 4c2 = c∗ 1 +2c1 (obtainedbyletting n =2in(3)).

(2)Ifwecombineformula g0 = c0 2(c2 + c0 h1 )+ c0 forthenumberof0-gaps in2DwiththedefinitionoftheEulercharacteristicwith0-adjacencyasan alternatingsumofthenumberof i-cellsin O , i =0, 1, 2,weobtain

(3)Oneofthefirstalgorithms[25]forthecomputationoftheEulercharacteristicofa2Ddigitalobject O countsthenumberofdifferentpixelconfigurationsinthe2 × 2blocksaroundthegridvertices.Theconfigurations Q1 , Q3 and QD thataffecttheEulercharacteristicarethosehavingexactlyone black(exactlyonewhite)pixel,withotherthreepixelsbeingwhite(black), andthosehavingexactlytwodiagonallyplacedblackpixels(andtwowhite ones),respectively.Theformertwoconfigurationscorrespondtoconvexand concaveverticesintheboundaryof O ,respectively,andthelatteronescorrespondto0-gaps,i.e., |QD | = g0 .TheEulercharacteristiciscomputedas follows:

0 (O )= 1 4 |Q1 | + 1 4 |Q3 |− 1 2 g0 ,

1 (O )= 1 4 |Q1 | + 1 4 |Q3 | + 1 2 g0

Fromthis,weget

andtheobviousrelation

Thesamerelationcanbeobtainedfrom

(c0 + g0 ) c1 + c2 (seeSect. 2.3).

(4)Usingtheformula[45]

weobtain

Thisisequivalenttothewell-knownexpressionoftheEulercharacteristic asthealternatingsumofthenumberof i-cellsinacomplex,i.e.,

5TwoAlternativeProofsofFormula(2)

Wegivetwonewshortproofsofformula(2),i.e.,oftheformula

Thefirstproofiscombinatorial,thesecondoneisbasedongraphtheory.Both proofsusetheinformationonthenumberofincidentcellsofvariousdimensions foragivencell,obtainedinastraightforwardmannerfromthecombinatorial coordinates,asdetailedinSect. 2.1.

(FirstProof)

Notethat c∗ n 2 isthenumberofboundary(n 2)-cellsthatmayormaynot be(n 2)-gaps.Thus,

, wherewedenotedas¯ g ∗ n 2 thenumberofnon-gapboundary(n 2)-cells(called alsototallyboundarycells[6, 11]ornubs[39]).

Letusconsiderthesetofallorderedpairs(f,e),where f isaboundary (n 1)-celland e isaboundary(n 2)-cell(non-gaporgap)incidentto f . Thereare c∗ n 1 cells f and

Eachboundary(n 1)cellisincidentto2(n 1)boundary(n 2)-cells [6, 11, 39].Eachboundarynon-gap(n 2)-cellisincidenttotwo,andeach(n 2)gapisincidenttofourboundary(n 1)-cells.Consequently,eachboundary (n 1)cellisin2(n 1),eachboundarynon-gap(n 2)-cellisintwoandeach (n 2)-gapisinfoursuchpairs(f,e).Thus

Theexistingproof[39]offormula(2)inessencereducestothisone.Ituses variousinterestingrelationsbetweendifferenttypesofcellsin Q obtainedby definingsomenewnotionsinthecubicalgrid.

(SecondProof)

Letusconsiderthegraph G =(N,A),wherethenodesin N correspond totheboundary(n 1)-cells.Therearetwotypesofarcsin A.Thefirsttype isassociatedwithnon-gapboundary(n 2)-cells,andthesecondisassociated with(n 2)-gaps.Eachnon-gapboundary(n 2)-cellisincidenttotwo,and each(n 2)-gapisincidenttofourboundary(n 1)-cells.Thetwoblackand twowhitevoxelsincidenttoan(n 2)-gap e intheblock Bn 2 (e)alternate cyclicallyaround e.Thus,twoofthefourboundary(n 1)-cellsincidentto e arebothincidenttooneofthetwowhitevoxels,andtheothertwoofthefour boundary(n 1)-cellsareincidenttotheotherwhitevoxel.

Weconnecttwonodesin N throughanarcin A ifthecorrespondingtwo boundary(n 1)-cellsshareacommonboundarynon-gap(n 2)-cell,orif theyshareacommon(n 2)-gapandtheyarebothincidenttothesamewhite voxel.Thus,eachnon-gapboundary(n 2)-cellgeneratesonearcin A,andeach (n 2)-gapgeneratestwoarcsin A.Eachboundary(n 1)-cellisconnectedto exactlyone(n 1)-cellacrosseachofits2(n 1)incident(n 2)-cells.

Thereare c∗ n 1 nodesin N and¯ g

2 arcsin A.Thedegreeofeach nodein N isequalto2(n 1)(thenumberofnon-gaporgap(n 2)-cellsin theboundaryofthe(n 1)-cellcorrespondingtothenode).

Thesumofdegreesofthenodesin N isequaltotwicethenumberofarcs in A,i.e.,

Asimilarideaofconstructingthedualgraphfromboundary2-and1-cells wasusedintheproofoftheformulainthespecialcase n =3[38].Thedifference isthatinourgraph G =(N,A),each(n 2)-gapcontributesonlytwo(andnot six,asinthe3Dcase)arcsto A,andeachnodein N hasdegree2(n 1) independentlyofwhetherthecorrespondingboundary(n 1)-cellisincidentto some(n 2)-gapsornot.

6ConclusionandFutureWork

Tworesearchfieldsintheareaoftopologicalimageanalysis,namelythecomputationofthenumberof0-gapsandthecomputationoftheEulercharacteristic of2Ddigitalobjects,arecloselyrelated.Thisisdemonstratedbyanexpression[9]thatlinksthenumberof0-gapsina2DdigitalobjectwithitsEuler characteristic.

Usingthisconnection,wegiveseveralnewformulaeforthenumberof0-gaps in2D,obtainedfromsomeknownformulaefortheEulercharacteristicofa2D digitalobject O [25, 43, 45].Alternatively,theformulaeforthenumberof0-gaps canbeusedtoobtainnewrelationsbetweentheEulercharacteristicandthe numberofcellsofvarioustypesintheassociatedcubicalcomplex.

Wegivetwonewshortandintuitiveproofsofone[39]ofthetwoknown formulaeforthenumberof(n 2)-gapsin nDdigitalobjects.Thetwoproofs relyontheinformationonthenumberofcellsintheboundaryandcoboundary ofagivencell,readilyobtainedfromthecellcoordinates(seeSect. 2.1).Combinatorialcoordinatesweredefinedrecentlyforsomenontraditional3Dgrids (suchasbodycenteredcubic,facecenteredcubicandthediamondgrids)[16–19],alternativetothetraditionalcubicone,andwereprovenusefulforstandard imageprocessingtaskssuchasboundaryextractionandcomputationofthe Eulercharacteristic[16, 34].

Thenumberof1-gapsin3Ddigitalobjectsiswell-understood,duetothe factthateachedgeexhibitinga1-gapisincidenttoexactlyfourboundaryfaces, andallotherboundaryedgesareincidenttoexactlytwosuchfaces.Thecomputationofthenumberof0-gapsin3Ddigitalobjectsdoesnotlenditselfto astraightforwardgeneralization,asverticesexhibitinga0-gaparenotuniquely characterizedbythenumberofincidentboundarycells.Itwouldbeinteresting toinvestigatetheconnectionbetweentheEulercharacteristicandthenumber of0-gapsin3Ddigitalobjects.

Acknowledgement. ThisworkhasbeenpartiallysupportedbytheMinistryofEducationandScienceoftheRepublicofSerbiawithintheProjectNo.34014.

References

1.Andres,E.,Acharya,R.,Sibata,C.H.:Discreteanalyticalhyperplanes.CVGIP: Graph.ModelImageProcess. 59(5),302–309(1997)

2.Bishnu,A.,Bhattacharya,B.B.,Kundu,M.K.,Murthy,C.A.,Acharya,T.:OnchipcomputationofEulernumberofabinaryimageforefficientdatabasesearch. In:Proceedingsofthe2001InternationalConferenceonImageProcessing,ICIP, pp.310–313(2001)

3.Bishnu,A.,Bhattacharya,B.B.,Kundu,M.K.,Murthy,C.A.,Acharya,T.:A pipelinearchitectureforcomputingtheEulernumberofabinaryimage.J.Syst. Archit. 51(8),470–487(2005)

4.Boutry,N.,G´eraud,T.,Najman,L.:HowtomakenDimageswell-composedwithoutinterpolation.In:2015IEEEInternationalConferenceonImageProcessing, ICIP2015,pp.2149–2153(2015)

5.Bribiesca,E.:ComputationoftheEulernumberusingthecontactperimeter.Comput.Math.Appl. 60(5),1364–1373(2010)

6.Brimkov,V.E.:Formulasforthenumberof(n 2)-gapsofbinaryobjectsinarbitrarydimension.DiscreteAppl.Math. 157(3),452–463(2009)

7.Brimkov,V.E.,Barneva,R.:Lineartimeconstant-workingspacealgorithmfor computingthegenusofadigitalobject.In:Bebis,G.,etal.(eds.)ISVC2008. LNCS,vol.5358,pp.669–677.Springer,Heidelberg(2008). https://doi.org/10. 1007/978-3-540-89639-5 64

8.Brimkov,V.E.,Klette,R.:Borderandsurfacetracing-theoreticalfoundations. IEEETrans.PatternAnal.Mach.Intell. 30(4),577–590(2008)

9.Brimkov,V.E.,Maimone,A.,Nordo,G.:Anexplicitformulaforthenumberof tunnelsindigitalobjects.CoRRabs/cs/0505084(2005). http://arxiv.org/abs/cs/ 0505084

10.Brimkov,V.E.,Maimone,A.,Nordo,G.:Countinggapsinbinarypictures.In: Reulke,R.,Eckardt,U.,Flach,B.,Knauer,U.,Polthier,K.(eds.)IWCIA2006. LNCS,vol.4040,pp.16–24.Springer,Heidelberg(2006). https://doi.org/10.1007/ 11774938 2

11.Brimkov,V.E.,Moroni,D.,Barneva,R.:Combinatorialrelationsfordigitalpictures.In:Kuba,A.,Ny´ul,L.G.,Pal´agyi,K.(eds.)DGCI2006.LNCS,vol.4245, pp.189–198.Springer,Heidelberg(2006). https://doi.org/10.1007/11907350 16

12.Brimkov,V.E.,Nordo,G.,Barneva,R.P.,Maimone,A.:Genusanddimensionof digitalimagesandtheirtime-andspace-efficientcomputation.Int.J.ShapeModel. 14(2),147–168(2008)

13.Chen,L.:Determiningthenumberofholesofa2Ddigitalcomponentiseasy. CoRRabs/1211.3812(2012)

14.Chen,M.,Yan,P.:AfastalgorithmtocalculatetheEulernumberforbinary images.PatternRecogn.Lett. 8(5),295–297(1988)

15.Cohen-Or,D.,Kaufman,A.E.:3Dlinevoxelizationandconnectivitycontrol.IEEE Comput.Graph.Appl. 17(6),80–87(1997)

16. Comi´c,L.,Magillo,P.:Repairing3DbinaryimagesusingtheBCCgridwitha 4-valuedcombinatorialcoordinatesystem.Inf.Sci.,toappear

17. Comi´c,L.,Nagy,B.:Atopologicalcoordinatesystemforthediamondcubicgrid. ActaCrystallogr.Sect.A 72(5),570–581(2016)

18. Comi´c,L.,Nagy,B.:Acombinatorialcoordinatesystemforthebody-centered cubicgrid.Graph.Models 87,11–22(2016)

19. Comi´c,L.,Nagy,B.:Atopological4-coordinatesystemforthefacecenteredcubic grid.PatternRecogn.Lett. 83,67–74(2016)

20.Dey,S.,Bhattacharya,B.B.,Kundu,M.K.,Acharya,T.:AFastalgorithmfor computingtheeulernumberofanimageanditsVLSIimplementation.In:13th InternationalConferenceonVLSIDesign(VLSIDesign2000),pp.330–335(2000)

21.Dey,S.,Bhattacharya,B.B.,Kundu,M.K.,Bishnu,A.,Acharya,T.:ACoprocessorforcomputingtheEulernumberofabinaryimageusingdivide-andconquerstrategy.Fundam.Inf. 76(1–2),75–89(2007)

22.D´ıaz-de-Le´onS.,J.L.,Sossa-Azuela,J.H.:OnthecomputationoftheEulernumber ofabinaryobject.PatternRecognit. 29(3),471–476(1996)

23.Dyer,C.R.:ComputingtheEulernumberofanimagefromItsquadtree.Comput. Graph.ImageProcess. 13,270–276(1980)

24.Fran¸con,J.,Schramm,J.-M.,Tajine,M.:Recognizingarithmeticstraightlines andplanes.In:Miguet,S.,Montanvert,A.,Ub´eda,S.(eds.)DGCI1996.LNCS, vol.1176,pp.139–150.Springer,Heidelberg(1996). https://doi.org/10.1007/3540-62005-2 12

25.Gray,S.:Localpropertiesofbinaryimagesintwodimensions.IEEETrans.Comput. 20,551–561(1971)

26.He,L.,Chao,Y.:AveryfastalgorithmforsimultaneouslyperformingconnectedcomponentlabelingandEulernumbercomputing.IEEETrans.ImageProcess. 24(9),2725–2735(2015)

27.He,L.,Chao,Y.,Suzuki,K.:Analgorithmforconnected-componentlabeling,hole labelingandEulernumbercomputing.J.Comput.Sci.Technol. 28(3),468–478 (2013)

28.He,L.,Yao,B.,Zhao,X.,Yang,Y.,Chao,Y.,Ohta,A.:Agraph-theory-based algorithmforEulernumbercomputing.IEICETrans. 98–D(2),457–461(2015)

29.Imiya,A.,Eckhardt,U.:TheEulercharacteristicofdiscreteobject.In:Ahronovitz, E.,Fiorio,C.(eds.)DGCI1997.LNCS,vol.1347,pp.161–174.Springer,Heidelberg (1997). https://doi.org/10.1007/BFb0024838

30.Kenmochi,Y.,Imiya,A.:Combinatorialtopologiesfordiscreteplanes.In:Nystr¨om, I.,SannitidiBaja,G.,Svensson,S.(eds.)DGCI2003.LNCS,vol.2886,pp.144–153.Springer,Heidelberg(2003). https://doi.org/10.1007/978-3-540-39966-7 13

31.Klette,R.,Rosenfeld,A.:DigitalGeometry.GeometricMethodsforDigitalPicture Analysis.MorganKaufmannPublishers,SanFrancisco(2004)

32.Kong,T.Y.,Rosenfeld,A.:Digitaltopology:introductionandsurvey.Comput. Vis.Graphi.ImageProcess. 48(3),357–393(1989)

33.Kovalevsky,V.A.:GeometryofLocallyFiniteSpaces(ComputerAgreeableTopologyandAlgorithmsforComputerImagery).EditingHouseDr.B¨arbelKovalevski, Berlin(2008)

34.Lachaud,J.-O.:Codingcellsofdigitalspaces:aframeworktowritegenericdigital topologyalgorithms.Electron.NotesDiscreteMath. 12,337–348(2003)

35.Latecki,L.J.:3Dwell-composedpictures.CVGIP:Graph.ModelImageProcess. 59(3),164–172(1997)

36.Latecki,L.J.,Eckhardt,U.,Rosenfeld,A.:Well-composedsets.Comput.Vis.Image Underst. 61(1),70–83(1995)

37.Lin,X.,Sha,Y.,Ji,J.,Wang,Y.:AproofofimageEulernumberformula.Sci. ChinaSer.F:Inf.Sci. 49(3),364–371(2006)

38.Maimone,A.,Nordo,G.:On1-gapsin3Ddigitalobjects.Filomat 22(3),85–91 (2011)

39.Maimone,A.,Nordo,G.:Aformulaforthenumberof(n 2)-gapsindigital n-objects.Filomat 27(4),547–557(2013)

40.Maimone,A.,Nordo,G.:Anoteondimensionandgapsindigitalgeometry.Filomat 31(5),1215–1227(2017)

41.Rosenfeld,A.,Kak,A.C.:DigitalPictureProcessing.AcademicPress,London (1982)

42.Sossa,H.:Onthenumberofholesofa2-Dbinaryobject.In:14thIAPRInternationalConferenceonMachineVisionApplications,MVA,pp.299–302(2015)

43.Sossa-Azuela,J.H.,Cuevas-Jim´enez,E.B.,Zaldivar-Navarro,D.:Alternativeway tocomputetheEulernumberofabinaryimage.J.Appl.Res.Technol. 9,335–341 (2011)

44.Sossa-Azuela,J.,Santiago-Montero,R.,P´erez-Cisneros,M.,Rubio-Espino,E.: ComputingtheEulernumberofabinaryimagebasedonavertexcodification. J.Appl.Res.Technol. 11(3),360–370(2013)

45.Sossa-Azuela,J.,Santiago-Montero,R.,P´erez-Cisneros,M.,Rubio-Espino,E.: AlternativeformulationstocomputethebinaryshapeEulernumber.IETComput. Vis. 8(3),171–181(2014)

46.Yagel,R.,Cohen,D.,Kaufman,A.E.:Discreteraytracing.IEEEComput.Graph. Appl. 12(5),19–28(1992)

47.Yao,B.,etal.:Anefficientstrategyforbit-quad-basedEulernumbercomputing algorithm.IEICETrans.Inf.Syst. E97.D(5),1374–1378(2014)

48.Zenzo,S.D.,Cinque,L.,Levialdi,S.:Run-basedalgorithmsforbinaryimageanalysisandprocessing.IEEETrans.PatternAnal.Mach.Intell. 18(1),83–89(1996)

49.Zhang,Z.,Moss,R.H.,Stoecker,W.V.:Anovelmorphologicaloperatortocalculate Eulernumber.In:MedicalImagingandAugmentedReality:FirstInternational Workshop,MIAR,pp.226–228(2001)

FixpointsofIteratedReductions withEquivalentDeletionRules

DepartmentofImageProcessingandComputerGraphics, UniversityofSzeged,Szeged,Hungary {palagyi,gnemeth}@inf.u-szeged.hu

Abstract. Areductiontransformsabinarypictureonlybydeleting someblackpointstowhiteones.Sequentialreductionstraversetheblack pointsofapicture,andfocusontheactuallyvisitedpointforpossibledeletion,whileparallelreductionsdeleteall‘deletable’blackpoints simultaneously.Tworeductionsarecalledequivalentiftheyproducethe sameresultforeachinputpicture.Adeletionruleissaidtobeequivalentifitprovidesapairofequivalentsequentialandparallelreductions. Thinningandshrinkingalgorithmsiteratereductionsuntilnopointsare deleted.Ifablackpointisnotdeletedinaniterationstep,itistaken intoconsiderationagaininthenextstep.Thisworkexaminefixpointsof iteratedreductionswithequivalentdeletionrules,i.e.,‘survival’points whoserecheckingisnotneededintheremainingiterations.

Keywords: Digitalgeometryandtopology Equivalentreduction Thinning

1Introduction

A digitalbinarypicture [5, 9]onagridisamappingthatassignsacolorof black or white toeachgridelementthatiscalleda point .A reduction [2]transformsabinarypictureonlybychangingsomeblackpointstowhiteones,which isreferredtoas deletion . Sequentialreduction sfocusonasingleblackpoint forpossibledeletionatatimeasitisillustratedinAlgorithm 1,while parallel reduction scandeleteallblackpointsfromapicturethatsatisfytheirdeletion rulessimultaneously,seeAlgorithm 2.Iteratedreductionsplayakeyroleinvarioustopologicalalgorithms,e.g. thinning [3, 7, 17](i.e.,alayer-by-layererosion forapproximating skeleton-likefeature sinatopology–preservingway)or reductiveshrinking [2](i.e.,thatiscapableofproducingaminimalstructurethatis topologicallyequivalenttotheoriginalobject).

Reductionsgenerallyclassifythesetofblackpoints B intotwosubsets: elementsofthe constraintset C (B ) ⊆ B maynotbedeleted,andthesetof interestingpoint s X = B \ C (B )areformedbythepotentiallydeletablepoints. Thedeletionrule R associatedwiththereductioninquestionisevaluatedonly fortheelementsin X ,andblackpointsin C (B )arenottakenintoconsideration.

c SpringerNatureSwitzerlandAG2018

R.P.Barnevaetal.(Eds.):IWCIA2018,LNCS11255,pp.17–27,2018. https://doi.org/10.1007/978-3-030-05288-1 2

Algorithm1. Sequentialreduction(thinning/shrinkingiterationstep)

Input:setofblackpoints B , constraintset C (B ), permutation Π ofset B \ C (B ),and deletionrule R

Output:setofblackpoints SB

//specifyingthesetofinterestingpoints X

X = B \ C (B )

//initializingtheset SB

SB = B

//traversalofset X accordingtopermutation Π foreach p ∈ X do

if R ( SB ∩ SR (p),C (B ) ∩ SR (p))= truethen //deletionofthesinglepoint p

SB = SB \{p}

Algorithm2. Parallelreduction(thinning/shrinkingiterationstep)

Input:setofblackpoints B , constraintset C (B ),and deletionrule R

Output:setofblackpoints PB

//specifyingthesetofinterestingpoints X

X = B \ C (B )

//specifyingthesetofdeletablepoints D

D = { p | p ∈ X and R ( B ∩ SR (p),C (B ) ∩ SR (p))= true }

//simultaneousdeletionofallpointsin D

PB = B \ D

Asequentialreductions(seeAlgorithm 1)requiresapermutationoftheinterestingpoints,sincetheproducedoutputmaydependontheorderinwhichthe pointsareselectedbythe foreach loop.Incontrast,theoutputpictureproduced byaparallelreductionisuniquelydeterminedbyAlgorithm 2.Forpracticalpurposesitisassumedthatallinputpicturesare finite (i.e.,theycontainfinitely manyblackpoints).

A deletionrule R isamappingthatassignsavalueof true or false toall possiblepointconfigurationsinthesupportof R.The support [3]of R appliedat apoint p isaminimalsetofpointswhosevaluesdeterminewhether p isdeleted by R.InAlgorithms 1 and 2,thesupportofdeletionrule R withrespectto aninterestingpoint p isdenotedby SR (p).Notethatallexistingthinningand shrinkingalgorithmsuselocalsupportswith‘small’diameters.

Aninterestingpoint p ∈ X is deletable bydeletionrule R fromthesetof points Y ⊆ B withtheconstraintset C (B ),if R(Y ∩ SR (p),C (B ) ∩ SR (p))= true. If R(Y ∩ SR (p),C (B ) ∩ SR (p))= false,thenpoint p iscalled non-deletable by R. Notethat Y = B inparallelreductions(seeAlgorithm 2),and Y = SB ⊆ B in

Another random document with no related content on Scribd:

The Project Gutenberg eBook of Ready, aye ready!

This ebook is for the use of anyone anywhere in the United States and most other parts of the world at no cost and with almost no restrictions whatsoever. You may copy it, give it away or re-use it under the terms of the Project Gutenberg License included with this ebook or online at www.gutenberg.org. If you are not located in the United States, you will have to check the laws of the country where you are located before using this eBook.

Title: Ready, aye ready!

Release date: May 3, 2024 [eBook #73528]

Language: English

Original publication: London: James Nisbet & Co., Limited, 1888

*** START OF THE PROJECT GUTENBERG EBOOK READY, AYE READY! ***

Transcriber's note: Unusual and inconsistent spelling is as printed.

READY, AYE READY!

AUTHOR OF "ST. AUSTIN'S LODGE," ETC.

London

JAMES NISBET & CO., LIMITED

22 BERNERS STREET, W.

CHAP.

At the Ballantyne Press, Edinburgh

CONTENTS.

I. A TOWN COTTAGE

II. DANGEROUS

III. THE NEIGHBOURS

IV. NANCY AND THE DOG

V. A BRAVE DEED

VI. SUNDAY MORNING

VII. GOING TO CHURCH

VIII. THE OLD MOTTO

IX. ARCHIE'S MOTHER

X. TEMPERS

XI. "MY ONLY BOY"

XII. WITH MISS WILMOT

XIII. MISSING

XIV. FALLEN BRICKS

XV. A PRIVATE TALK

XVI. MRS. MASON'S OPINIONS

XVII. PEOPLE AND THINGS

XVIII. THE GARDINERS

XIX. THAT GIRL BESS!

XX. TROUBLE

XXI. THE GREAT DREAD

XXII. LIFE LESSONS

XXIII. GOOD ADVICE

XXIV. THE RECTOR'S WISH

XXV. REST

XXVI. A LOSS

XXVII. HOW THINGS CAME ABOUT

READY, AYE READY!

CHAPTER I.

A TOWN COTTAGE.

"AND they call this 'Woodbine Cottage'!" said Susan Dunn, gazing out upon a paved street, where several unwashed children were disporting themselves in a gutter.

She looked dusty and unhappy, and she spoke disconsolately,—all three things rare with Susan Dunn, for a neater and brighter and more smiling little roundabout of a woman could in general be nowhere found.

The cottage wore an air of confusion. A damp floor told of recent scrubbing, as also did pail and soap, not to speak of Mrs. Dunn's own rolled-up sleeves and bare arms.

In one corner stood a pile of chairs: in another a small wooden cupboard. An empty book-case lay on the table, and through the open door might be seen a folded carpet, waiting for admission.

Side by side with the book-case, on the round table, sat cross-legged a little girl, about seven years old, contemplating the scene with sober eyes. She had plainly climbed there to be out of the way.

There had been a family flitting from an old home to a new. No unusual event this, in the lives of many men and

women, but very unusual in the lives of Richard and Susan Dunn.

For during more than fifteen years Dunn had worked as a regularly attached labourer, under Messrs. Horry, Builders, &c.; and not merely as a labourer, but as a clever and skilful "handy man," though, unhappily for himself, not as a skilled artisan. A widowed mother had been unable to afford the seven years' apprenticeship in those days requisite; and Richard Dunn's abilities were thereby handicapped through his after life.

For fifteen years they had known no change; and now suddenly a change had become necessary. Times were bad for business men, and Dunn's employers had failed. Thereby many working-men were cast adrift, and compelled to flit elsewhere in search of employment.

Dunn found what he needed, with less difficulty than he had expected. But he had to quit the pretty cottage and well-cultivated garden-slip which had long been the pride of his heart; to say good-bye to friends of many years' standing; and to find himself fresh quarters at Littleburgh.

"Woodbine Cottage" had sounded hopeful, when first he was advised to take a look at it. Dunn soon found, however, that dreams of country prettiness and twining creepers must be put aside.

Littleburgh was a bustling manufacturing town, of perhaps some nine or ten thousand inhabitants, and, viewed from a money point of view, it might be regarded as a very thriving place. There were cloth-factories and other factories, some of the former being worked almost entirely by women. Fresh houses and streets were being run up in

all directions: so the builders were just then having a fairly good time of it. Of course rents were proportionately dear.

The town lay in a flat unbeautiful neighbourhood; very different from the fair and hilly landscape the Dunns were used to look upon. Long rows of small red or white houses, as much alike one to another as a supply of pill-boxes, stretched away to the east and south; and beyond them lay wide brick-fields, with a kiln here and there.

"Woodbine Cottage" stood exactly in the middle of one such row. It was flanked by "Rose Cottage" and "Myrtle Cottage." But no roses, myrtles, or woodbine grew anywhere near.

The cottage contained four rooms, two below and two above; so it did well enough for a small family as to space. It opened straight upon the pavement, without an inch of garden in front, and with only a minute yard behind. No doubt the Dunns were well off to have so much. Some cottages in Littleburgh could boast no back yard at all. Still —when they should think of the garden they had left, the pinks and sweet-williams, the roses and geraniums, cannot we imagine how they would feel?

They had arrived quite early that morning, and Dunn had immediately gone straight off to his new work. He was not a man to waste one day without necessity.

For hours since, Susan and her eldest girl, Nancy, had been scrubbing and scouring. The bedrooms and the kitchen were now almost in order. Only the little parlour remained. Susan was bent upon getting that straight too, before her husband should come in.

"Just to make him feel home-like," she said once or twice.

But Susan was very tired, and her back ached. They had started early by train, and she had been up late the night before. So at length she stood still for a minute to rest, gazing out upon the street, and then it was that the words escaped her lips,—

"They call this 'Woodbine Cottage!'"

"It isn't so pretty as home, mother," a sedate small voice said, and the little maiden on the table folded her hands.

"Pretty! There isn't a scrap of woodbine anywhere," said Susan.

"Won't father get some?"

"It wouldn't grow here, if he did, I expect."

"Won't father get some nice garden-mould, mother?"

Susan shook her head. "I don't know what he'll do yet, child. We've got to make the best of things—somehow."

Then she fell to her work anew. The boards were getting dry, and soon it was time to lay the carpet down. The elder girl, Nancy, was upstairs, finishing the bedrooms; and the boy, Dick, had gone to assist. So Susie climbed off the table, and offered her tiny help. It was little, if anything, that she could really do; yet Susan would not spurn that little or nothing. For the child wanted to be useful, and that was good both for child and mother. So Susan had often said.

"Pull harder, Susie. It is not straight yet," Susan said encouragingly.

All at once a stronger hand grasped Susie's corner, and in a trice the matter was accomplished. Susan said, "Why, Richard!" and looked up.

No, not Richard Dunn. A stout motherly woman was raising herself erect.

"That was what you wanted, wasn't it?" she said. "So you've only just come?"

Susan answered civilly but coldly. She did not wish to plunge into fresh acquaintances, without knowing something first of the people around. Like a wise mother, she had a great notion of choosing good and desirable companions for her children.

But the new-comer, smiling good-humouredly, seemed unconscious of her coldness.

"Now, I suppose you'll put the cupboard into that recess. It'll just fit there, if I'm not mistaken. What a handsome bit of furniture it is! And that clock has the look of something uncommon. Old, isn't it?"

"Yes,—it came from Holland. It's more than a hundred years old," said Susan.

"Wouldn't Mr. Rawdon give something to have it! He's a wonderful fancier of old furniture and the like!"

Mr. Rawdon was Richard Dunn's new employer; but Susan said nothing. She could hardly refuse, however, the kindly offered help, and for a few minutes the two worked hard.

"That looks well!" the other said presently. "Is this little one your only child? Pretty," was murmured in a lower key.

"No, I've two others."

"Older or younger?"

"Dick is one year older; and Nancy is sixteen."

"Sixteen!"

The stout woman paused, looked at Susan, looked out of the window, and then once more looked back at Susan.

"I don't know you yet, my dear," she said. "I don't know whether you're one that minds a word of advice. I shouldn't wonder if you'll count it an Impertinent interference. Maybe I should in your place. For I'm a stranger to you,—and if I say I'm Mrs. Mason from over the way, you'll not be much the wiser. But there's one thing I should like to say to you, now you've just come here, and don't know the ways of the place nor the people,—if you'll not take it ill, that's to say?"

Susan simply answered,—"No."

Mrs. Mason's face broke into a smile.

"I like that," she said. "I like a woman that don't pour out a rush of words, like water gallopading out of a spout. 'No' was enough: and you're right to say just that, and not a word more."

Susan privately thought that Mrs. Mason was admiring a virtue to which she had not herself attained.

"Well, but what I was going to say," resumed the other, "what I was going to say to you, was this: If you can help it, don't you let that girl of yours work in the factories!"

"My Nancy work in the factory!" Susan's comely face lighted up with a hot flush. "My Nancy! Thank you, Mrs. Mason. No, I'm not angry, indeed—but it's such an idea as never came into my head before!"

Susan stopped, and suddenly cooled down.

"But there! You don't know my Nannie yet," she said quietly.

"No, I don't, Mrs. Dunn."

"Nor her father—nor me!"

Mrs. Mason's face broke into a genial smile.

"I think I'm getting to know you fast, Mrs. Dunn," she said, "and the faster the better. I'm sorry I said a word about the factories, as it seems to put you out,—but, dear me, if you knew what I know! The young mothers that are out all day, working, and leaving their little ones with none to care for them! And the girls of fifteen and sixteen, who do for themselves, and won't brook a word of control from father nor mother nor anybody! Not but what there's good girls among them, and good women too. But I always do say there's a lot of temptation, specially for young things. And if I was you, I wouldn't send a girl of mine there without I was downright obliged. I wouldn't, Mrs. Dunn!"

RICHARD DUNN was walking briskly homeward, half-anhour later, with a young man by his side.

He did not look depressed, though the change in his home and circumstances had been to him, as to Susan, no slight trouble. Dunn was a man of cheerful spirit; and the cheerfulness could be seen in his very air. He carried himself with an alert vigour, usually more characteristic of the artisan than of the labourer. In truth, Dunn occupied a position decidedly superior to that of the ordinary labourer, though of course inferior to that of the journeyman. He had a peculiar knack of picking up knowledge in all directions, and was actually capable of "handling the tools" in an emergency, though not usually permitted to exercise this feat in the trade.

Besides being thoroughly "handy" in his ordinary work, Dunn was something of a scholar. He read much, and thought over what he read. But Dunn was the last man ever to make a boast of these acquirements. People were often long in finding out his intelligence and capabilities.

Not always, however. For the young fellow by his side, Archibald Stuart by name, a last-year apprentice in the workshop of Mr. Rawdon, had already scented out "something uncommon" in the new workman. Dunn had known Archibald Stuart's father many years before, and this fact drew the two together. Moreover, Archibald's homeward road lay past Woodbine Cottage. Halfway thither, he overtook Dunn, and remained with him.

"You'll tell your mother we're here," Dunn said, as they drew near the end of their walk. "Dear me, I remember her well as a fine tall lass! There wasn't a handsomer girl about,

nor one that held her head higher. Your father did think a deal of her, to be sure."

"She has been a good mother to me," Archibald said. He was a fine tall young fellow himself, doubtless taking after his mother.

"And you're her only one? No brothers nor sisters?"

"None," Archibald said. "Only mother and me."

Dunn stood still. "I won't ask you in to-day," he said. "We're all in a mess—only come this morning. But we shall see you here soon; and you: mother too."

"I hope so," Archibald answered. He did not speak with certainty, and a slight shade crossed his face.

Almost instantly the shadow passed, and was replaced by a look of admiration. For the door opened, and a voice said, "Father!" as a young girl came out to welcome him.

The young girl was not alone. A plump little woman, all smiles, stood close behind, and two children also.

Archibald saw only Nancy, however. He was oblivious to the presence of anything or anybody else; and he stood stock-still, gazing hard, with never a thought of how his conduct might seem to others.

I do not think his admiration of Nancy could be wondered at,—she was such a very pretty girl: not only young and fresh-complexioned, but really pretty, almost beautiful. Her features were regular, and her rosy month and blue eyes seemed only made to smile. Besides all this, Nancy had a nice figure, and a soft winning manner, and hair and dress and hands all most daintily neat. She wore a

print dress, made by herself; but the dress fitted like a glove; and her linen collar and cuffs were spotless. So it was not surprising that young Stuart stood as if he had been moonstruck, almost forgetting where he was.

"Father," Nancy said again, "it's all so nice. Come in and see how we've put things. I am sure you are tired."

Archibald thought he had never in his life heard so kind and sweet a voice before. But nobody noticed him. Dunn had responded to the petition, going straight in. Archibald could see two children clinging to him, and the wife's hand on his arm. He saw all that without caring particularly. What did strike him was the upturned girlish face, with loving eyes and rosy lips, and the voice which he could still hear repeating,—

"Come and see it all, father! We have been so hard at work all day, and now things do really look like home."

Then Archibald suddenly woke up to the fact that he had no business there. He strode fast off to his own home; and all the way thither, he could not get Nancy Dunn's face out of his mind.

Mrs. Stuart lived two streets farther on. The cottages here were rather smaller than Woodbine Cottage. It was now nearly eight years since the death of Archibald's father, and the widow had had a hard time of it since. The wish of Mrs. Stuart's heart had been that her boy Archie should be in his father's trade. All her powers were bent to this aim. Through the long years of his apprenticeship she had pinched and denied herself in every possible way for his sake. She could not endure that her boy should be one whit less respectably dressed than his fellow-apprentices. He always had his best and second best as well as his working

suits. Nay, even in the matter of his pocket-money, she would not keep him shorter than others.

It may be doubted whether Mrs. Stuart was quite wise here. Archie would have been none the worse for some necessary self-denial in daily life.

But Mrs. Stuart held her head high still, as she had done in younger days. She held it high for her boy even more than for herself. She was very particular as to whom she would know and very anxious about Archie's acquaintances.

Six years and a half of the apprenticeship were over. Six months more, and Archie would be a full-blown artisan, receiving man's pay. Mrs. Stuart's hope and expectation were that then he would repay her long toil and self-denial. She did not intend him to marry early; and when he did marry, her daughter-in-law was to be unexceptionable in character and position. Mrs. Stuart, daughter of a farmer, widow of an artisan, would be content with nothing lower in the social scale than her own standing for this only son.

Yes, certainly Mrs. Stuart held her head high; everybody said so. She held it high literally as well as figuratively, being very tall and thin and upright. A greater contrast could hardly have been found than between this coldmannered handsome woman, and the bright little roundabout Susan Dunn.

Even towards Archibald, Mrs. Stuart was not gentle or warm in manner. Dearly, even passionately, as she loved her boy, he seldom had from her a look or word of tenderness. She indulged him in many ways, but she expected him always to conform to her will. As yet they had

seldom had collisions. Archie was a good-humoured kindhearted fellow.

"Mother," said he, about an hour after his return, "there's a family of Dunns come here to live. Dunn spoke to me to-day, and said he used to know you and father."

Mrs. Stuart did not seem very much interested. Archibald had to make the remark over again before it brought a response.

"Dunn!" she said indifferently. "Yes, I remember a man of that name. He used to work under your father about the time we were married."

"He seems a nice sort of fellow," Archie observed.

"I never cared much about him. He never came in my way," said Mrs. Stuart.

"But I think they are people we shall like," persisted Archie. "And they don't know anybody here; you might befriend them, I should think. Couldn't you?"

"I dare say I could if I chose," said Mrs. Stuart stiffly.

"And you'll choose, won't you—if I ask it?" Archie put on a coaxing air, sometimes effectual.

"What makes you so keen after them, all in a hurry?" demanded Mrs. Stuart. "That isn't your fashion commonly."

Archie reddened, feeling conscious. He did not want to speak about Nancy.

"I like Dunn," he said. "We walked home together, part of the way. He's a scholar, and no mistake. Got lots of

books, and read them all. I found out by asking—something he said about father started me off. And he wouldn't mind lending me a volume now and then. I should like that! Why, I should think he reads more books in a year than all the other men in our works put together."

"Mayn't be any better man for that," said Mrs. Stuart snappishly. "I don't hold with such a lot of reading. Has he got any family?"

Archie could only answer, "Yes."

"How many children?"

"Three, I believe."

"Any grown-up girls?"

"Only one," said Archie.

"You've not seen her?"

"Only just a glimpse in passing," said Archie, with a careless air.

Mrs. Stuart's mouth fell into a grim set, and she worked fast.

"I didn't speak to any of them except Dunn," pursued Archie. "He's a first-rate fellow, mother, really. I wish you'd get to know them."

No answer. Mrs. Stuart sewed on resolutely, clicking her needle fiercely at every stitch, in a manner peculiar to herself. It boded temper.

Archie felt irritated, but he knew well that it was wiser to say no more just then. Silence reigned for some minutes.

Turn static files into dynamic content formats.

Create a flipbook
Issuu converts static files into: digital portfolios, online yearbooks, online catalogs, digital photo albums and more. Sign up and create your flipbook.