Articulated Motion and Deformable Objects: Second International Workshop, AMDO 2002 Palma de Mallorca, Spain, November 21–23, 2002 Proceedings

266 Pages·2002·7.25 MB·English

by  Kevin T. McDonnell

Lecture Notes in Computer Science 2492
EditedbyG.Goos,J.Hartmanis,andJ.vanLeeuwen
3
Berlin
Heidelberg
NewYork
Barcelona
HongKong
London
Milan
Paris
Tokyo
Francisco Jose´ Perales
Edwin R. Hancock (Eds.)
Articulated Motion
and Deformable Objects
Second International Workshop, AMDO 2002
PalmadeMallorca,Spain,November21-23,2002
Proceedings
1 3
SeriesEditors
GerhardGoos,KarlsruheUniversity,Germany
JurisHartmanis,CornellUniversity,NY,USA
JanvanLeeuwen,UtrechtUniversity,TheNetherlands
VolumeEditors
FranciscoJose´Perales
Director,ComputerGraphicsandVisionGroup
Sub-DirectorComputerScienceEngineeringSchool(EPS-UIB)
DepartmentofComputerScienceandMaths
C/Valldemossakm7.5,PC07071,PalmadeMallorca,Baleares,Spain
E-mail:[email protected]
EdwinR.Hancock
DepartmentofComputerScience,UniversityofYork
YO15DDYork,UK
E-mail:[email protected]
Cataloging-in-PublicationDataappliedfor
AcatalogrecordforthisbookisavailablefromtheLibraryofCongress.
BibliographicinformationpublishedbyDieDeutscheBibliothek
DieDeutscheBibliothekliststhispublicationintheDeutscheNationalbibliografie;
detailedbibliographicdataisavailableintheInternetathttp://dnb.ddb.de
CRSubjectClassification(1998):I.4,I.2.10,I.6,I.5,I.3
ISSN0302-9743
ISBN3-540-00149-2Springer-VerlagBerlinHeidelbergNewYork
Thisworkissubjecttocopyright.Allrightsarereserved,whetherthewholeorpartofthematerialis
concerned,specificallytherightsoftranslation,reprinting,re-useofillustrations,recitation,broadcasting,
reproductiononmicrofilmsorinanyotherway,andstorageindatabanks.Duplicationofthispublication
orpartsthereofispermittedonlyundertheprovisionsoftheGermanCopyrightLawofSeptember9,1965,
initscurrentversion,andpermissionforusemustalwaysbeobtainedfromSpringer-Verlag.Violationsare
liableforprosecutionundertheGermanCopyrightLaw.
Springer-VerlagBerlinHeidelbergNewYork
amemberofBertelsmannSpringerScience+BusinessMediaGmbH
http://www.springer.de
©Springer-VerlagBerlinHeidelberg2002
PrintedinGermany
Typesetting:Camera-readybyauthor,dataconversionbyOlgunComputergrafik
Printedonacid-freepaper SPIN:10870677 06/3142 543210
Preface
The AMDO 2002 workshop took place at the Universitat de les Illes Balears
(UIB)on21–23November,2002,sponsoredbytheInternationalAssociationfor
Pattern Recognition (IAPR), AERFAI (Spanish Association for Pattern Recog-
nitionandImageAnalysis),theCICYT(ComisionInterministerialdeCienciay
Tecnologia,SpanishGovernment),theEuropeanCommission(HumanPotential
Programme:HighLevelScientificConferences),andtheMathematicsandCom-
puter Science Department of the UIB. In addition, Egse (Eurographics Spanish
Chapter)helpedwithdisseminationandpromotion.Thesubjectoftheworkshop
wasongoingresearchinarticulatedmotiononasequenceofimagesandsophisti-
cated models for deformable objects. The goals of these areas are to understand
and interpret the motion of complex objects that can be found in sequences
of images in the real world. These topics (geometric and physically deformable
models, motion analysis, articulated models and animation, visualization of de-
formable models, 3D recovery from motion, single or multiple human motion
analysis and synthesis, applications of deformable models and motion analysis,
face tracking, recovering and recognition models, etc.) are actually very inter-
esting ways of organizing important research to solve more general problems.
Anotherobjectiveofthisworkshopwastorelatefieldsusingcomputergraphics,
computer animation, or applications in several disciplines combining synthetic
andanalyticalimages.Inthisregarditisofparticularinteresttoencouragelinks
betweenresearchersinareasofcomputervisionandcomputergraphicswhohave
common problems and frequently use similar techniques.
Theworkshopincludedfivesessionsofpresentedpapersandtwotutorials.In-
vitedtalkstreatingvariousaspectsofthetopicswere:Haptic-basedDeformable
Solid of Arbitrary Topology, by Prof. H. Qin from the State University of New
York at Stony Brook (USA); Passive 3D Human Motion Capture: Vision-Based
TrackingMeetsComputerAnimation,byProf.I.KakadiarisfromHoustonUni-
versity(USA);andRecoveringNon-rigid3DShapeUsingaPlane+ParallaxAp-
proach, by Prof. N. Perez de la Blanca from the University of Granada (Spain).
November 2002 F.J. Perales, E.R. Hancock
Program Co-Chairs
Organization
AMDO 2002 was organized by the Department of Mathematics and Computer
Science, Universitat de les Illes Balears (UIB) in cooperation with IAPR (Inter-
nationalAssociationforPatternRecognition)andAERFAI(SpanishAssociation
for Pattern Recognition and Image Analysis).
Program Committee
General Workshop Co-Chairs: F.J.Perales,MathematicsandComputerScience
Department, UIB (Spain)
E.R.Hancock,DepartmentofComputerScience,
University of York (UK)
Organizing Chairs: M.Gonzalez,A.Igelmo,R.Mas,P.M.Mascaro´,
A. Mir, P. Palmer, F.J. Perales, J.M. Buades,
UIB (Spain)
Tutorial Chairs: M. Gonzalez, R. Mas, A. Mir, F.J. Perales, UIB
(Spain)
VIII Organization
Programme Committee
Aloimonos, Y. University of Maryland, USA
Aggarwal, J.K. University of Texas, USA
Ayache, N. INRIA, Sophia-Antipolis, France
Badler, N.I. University of Pennsylvania, USA
Boulic, R. EPFL, Switzerland
Brunet, P. University Polytechnic Catalonia, Spain
Cipolla, R. University of Cambrige, UK
Cohen, I. University of Southern California, USA
Davis, L.S. University of Maryland, USA
Del Bimbo, A. University di Firenze, Italy
Gong, S. Queen Mary and Westfield Coll., UK
Kakadiaris, I.A. University of Houston, USA
Kittler, J. University of Surrey, UK
Kunii, T.L. University of Hosei, Japan
Huang, T.S. University of Urbana-Champaign, USA
Jain, A. Michigan State University, USA
Medioni, G. University of Southern California, USA
Metaxas, D. University of Pennsylvania, USA
Nagel, H.-H. Institut fu¨r Algorithmen und Kognitive Systeme,
Germany
Nastar, C. LUV, France
Pentland, A. Media Lab, MIT, USA
Perez de la Blanca, N. University of Granada, Spain
Qin, H. Stony Brook University, New York, USA
Sanfeliu, A. IRI, CSIC-UPC, Spain
Seron, F. University of Zaragoza, Spain
Shirai, Y. University of Osaka, Japan
Susin, A. University Polytechnic Catalonia, Spain
Terzopoulos, D. University of Toronto, Canada
Teixeira, J.C. FCTUC, Portugal
Thalmann, D. EPFL, Switzerland
Villanueva, J. UAB-CVC, Spain
Sponsoring Institutions
IAPR (International Association for Pattern Recognition)
European Commission (Human Potential Programme)
MCyT (Ministerio de Ciencia y Tecnologia, Spanish Government)
TIC2001-5233-E
MathematicsandComputerScienceDepartment,UniversitatdelesIllesBalears
(UIB)
Conselleria d’Innovacio´ i Energia (Govern de les Illes Balears)
AERFAI (Spanish Association for Pattern Recognition and Image Analysis)
Table of Contents
ArticulatedMotionandDeformableObjectsAMDO2002
Virtual Clay: Haptics-Based Deformable Solids of Arbitrary Topology .... 1
Kevin T. McDonnell and Hong Qin
g-HDAF Multiresolution Deformable Models .......................... 21
Ioannis A. Kakadiaris, Emmanuel Papadakis, Lixin Shen,
Donald Kouri, and David Hoffman
Muscle-Driven Motion Simulation Based on Deformable Human Model
Constructed from Real Anatomical Slice Data ......................... 32
Hiroshi Inaba, Shin-ya Miyazaki, and Jun-ichi Hasegawa
Model Acquisition Using Shape-from-Shading.......................... 43
Antonio Robles-Kelly and Edwin R. Hancock
A Computational Algebraic Topology Model
for the Deformation of Curves ....................................... 56
M.F. Auclair-Fortier, P. Poulin, D. Ziou, and M. Allili
P3DMA: A Physical 3D Deformable Modelling and Animation System.... 68
Miquel Mascaro´ Portells, Arnau Mir, and Francisco Perales
A Novel Approach to Generate Multiple Shape Models
for Tracking Applications ........................................... 80
Daniel Ponsa and F. Xavier Roca
Real-Time Human Motion Analysis Based on Analysis
of Silhouette Contour and Color Blob................................. 92
Ryuya Hoshino, Daisaku Arita, Satoshi Yonemoto,
and Rin-ichiro Taniguchi
Human Body Model Acquisition and Motion Capture Using Voxel Data .. 104
Ivana Miki´c, Mohan Trivedi, Edward Hunter, and Pamela Cosman
3D Body Reconstruction for Immersive Interaction ..................... 119
Isaac Cohen and Mun Wai Lee
Wide-Range Tracking of Hands in Real-Time.......................... 131
Yoshio Iwai, Tomohiro Mashita, and Masahiko Yachida
Recognition, Tracking, and Reconstruction of Human Motion............ 142
J. Sullivan, M. Eriksson, and S. Carlsson
X Table of Contents
Tracking the Human Body Using Multiple Predictors................... 155
Rui M. Jesus, Arnaldo J. Abrantes, and Jorge S. Marques
Motion Estimation of Articulated Objects from Perspective Views ....... 165
Xiaoyun Zhang, Yuncai Liu, and Thomas S. Huang
Gesture and Posture Estimation by Using Locally Linear Regression...... 177
Yoshio Iwai, Keita Manjoh, and Masahiko Yachida
aSpaces: Action Spaces for Recognition and Synthesis of Human Actions .. 189
Jordi Gonza`lez, Javier Varona, F. Xavier Roca,
and Juan Jos´e Villanueva
Face Recognition Based on Efficient Facial Scale Estimation............. 201
Takatsugu Hirayama, Yoshio Iwai, and Masahiko Yachida
Eyebrow Movement Analysis over Real-Time Video Sequences
for Synthetic Representation ........................................ 213
Ana C. Andr´es del Valle and Jean-Luc Dugelay
Software Laboratory for Physical Based Human Body Animation ........ 226
Francisco Rojas, Sandra Baldassarri, and Francisco J. Sero´n
Computer Visual System Analyzing the Influence of Stimulants
on Human Motion ................................................. 241
Ryszard S. Choras and Michal Choras
Recovering Non-rigid 3D Shape Using a Plane+Parallax Approach ....... 251
Nicola´s P´erez de la Blanca and Antonio Garrido
Author Index ................................................. 257
Virtual Clay: Haptics-Based Deformable Solids
of Arbitrary Topology
Kevin T. McDonnell and Hong Qin
Department of Computer Science
State University of New York at Stony Brook
Stony Brook, NY 11794–4400
{ktm,qin}@cs.sunysb.edu
Abstract. ThispaperpresentsVirtualClayasanovel,interactive,dy-
namic,haptics-baseddeformablesolidofarbitrarytopology.OurVirtual
Claymethodologyisaunique,powerfulvisualmodelingparadigmwhich
isfoundedupontheintegrationof(1)deformablemodels,(2)free-form,
spline-based solids, (3) procedural subdivision solids of arbitrary topol-
ogy,and(4)dynamicobjectsgovernedbyphysicallaws.Solidgeometry
exhibits much greater modeling potential and superior advantages to
popular surface-based techniques in visual computing. This is primarily
because a CAD-based solid representation of a real-world physical ob-
ject is both geometrically accurate and topologically unambiguous. We
first introduce the concept of Virtual Clay based on dynamic subdivi-
sionsolids.Then,weformulatethemathematicsofVirtualClaythrough
the integration of the geometry of subdivision solids with the principle
of physics-based CAGD. Our Virtual Clay models respond to applied
forces in a natural and predictive manner and offer the user the illusion
ofmanipulatingsemi-elasticclayintherealworld.Weshowcaseexample
sculptures created with our Virtual Clay sculpting environment, which
isequippedwithalargevarietyofreal-time,intuitivesculptingtoolkits.
TheversatilityofourVirtualClaytechniquesallowsuserstomodifythe
topologyofsculptedobjectseasily,whiletheinherentphysicalproperties
areexploitedtoprovideanaturalinterfacefordirect,force-baseddefor-
mation.Moreimportantly,oursculptingsystemsupportsnaturalhaptic
interactiontoprovidetheuserwitharealisticsculptingexperience.Itis
our hope that our Virtual Clay graphics system can become a powerful
tool in graphics, computer vision, animation, computer art, interactive
techniques, and virtual environments.
1 Introduction and Rationale
Existing solid modeling techniques are oftentimes based on one or several of the
following geometric foundations: implicit surfaces, including Constructive Solid
Geometry (CSG) and blobby models; Boundary representations (B-reps), such
as subdivision surfaces and general polygonal meshes; spline-based solids, in-
cluding B´ezier and B-spline solids; and cell decompositions, such as voxel-based
models.Althoughtheseapproachesareidealforcertainapplications,eachtends
F.J.PeralesandE.R.Hancock(Eds.):AMDO2002,LNCS2492,pp.1–20,2002.
(cid:1)c Springer-VerlagBerlinHeidelberg2002