文档下载 文档下载导航
设为首页 | 加入收藏
搜索 请输入内容:  
 导航当前位置: 文档下载 > 所有分类 > Application of the Total Least Squares Procedure to Linear View Interpolation
侵权投诉

Application of the Total Least Squares Procedure to Linear View Interpolation

It is shown that, in comparison to the results obtained from a conventional least squares approach, a total least squares solution leads to significant improvements in the geometry and appearance of images synthesised in a linear combination of views proce

ApplicationoftheTotalLeast

SquaresProceduretoLinearView

Interpolation

DMKennedy,2BFBuxtonand1JHGilby1SiraLimited,SouthHill,Chislehurst,KentBR75EH2DepartmentofComputerScience,UniversityCollege

London,GowerStreet,LondonWC1E6BT

dawn.kennedy@ptp.sira.co.uk,b.buxton@cs.ucl.ac.uk,john.gilby@sira.co.uk1

Abstract

Itisshownthat,incomparisontotheresultsobtainedfromaconventionalleastsquaresapproach,atotalleastsquaressolutionleadstosignificantimprovementsinthegeometryandappearanceofimagessynthesisedinalinearcombinationofviewsprocedure.Useofthetotalleastsquarescriterionisappropriatewhenerrorsonthecontrolpointsareindependentlyandidenticallydistributedbetweenthebasisimagesandthetargetimagebeingsynthesised.Whenthisisnotbethecaseitispointedoutthatthegeneralisedtotalleastsquaresprocedureshouldbeused.Asyntheticobjectisusedtoevaluatetheimprovementingeometricaccuracyobtainedbyuseofthetotalleastsquaressolutionincomparisontoaclassicalleastsquaresmethod.Simulatedandrealimagesoflaboratorytestobjectsaresimilarlyusedtoillustratetheimprovementinappearanceofimagesreconstructedbymeansofthetotalleastsquaresprocedure.

1Introduction

Theproblemofviewsynthesis,inwhichnovelviewsofobjectsandscenesareconstructedfromasmallsetofbasisviews,hasreceivedmuchattentionrecently[3-11].Usually,butnotinvariably,thebasisviewsareexistingimagesandthereconstructionisbasedonaformofimageinterpolationormorphingthatoriginatedwiththeworkofUllmanandBasrionobjectrecognition.Theyfirstshowed[1]thatitispossible,toagoodapproximation,toderivelinearrelationshipsbetweenthegeometryofthreeviewsofascene.Sincethenthetheoryhasbeengreatlyextended,inparticularbytheapplicationofgeometricinvarianceprinciples[seeforexample12and13]anddevelopmentofthetriandquadrifocaltensors[14-16],butitappearsthatasubtlepoint,firstnoticedrecentlyincomputervisioninthecontextofestimatingthefundamentalmatrix[17],thewayimagemeasurementerrorsaretreatedhasnotbeenaddressed.Theaimofthispaperistoshowhow,fortheapproximatelinearmorphingrelationshipsmostfrequentlyusedinpractice,acorrecterroranalysismaybecarriedoutandtoevaluatetheresultingimprovementsinthegeometryandappearanceofthereconstructed

Application of the Total Least Squares Procedure to Linear View Interpolation

images.

305

第1页

你可能喜欢

热门文档

相关文档

站点地图 | 文档上传 | 侵权投诉 | 手机版
本站所有资源均来自互联网,本站只负责收集和整理,均不承担任何法律责任,如有侵权等其它行为请联系我们.
文档下载 Copyright 2013 doc.guandang.net All Rights Reserved.  -  email