Second International Conference on

3-D Imaging and Modeling

October 4-8, 1999

Château Laurier Hotel, Ottawa, Canada

Monday, October 4

7:30-18:00    Registration

09:00-12:00  Tutorial Sessions

Session #1      Active 3D Sensing
                     J. Angelo Beraldin and François Blais, National Research Council Canada

This tutorial provides an introduction to some of the most popular active 3D sensing techniques found in the literature and in commercial products. Although the emphasis of the course is focused on active triangulation and imaging time of flight systems, other techniques such as optical interferometry and passive methods are briefly discussed.

For each group of 3D method, optical, hardware, and algorithmic related topics as well as case studies are discussed. Basic knowledge of 3D measurement principle, Gaussian beam propagation and optical signal detection with solid-state detectors is provided so that critical aspects like image resolution and range image precision can be evaluated. Furthermore, algorithms related to signal extraction, control of scanning mechanisms, calibration, and some specific case studies are presented.

Benefits

 The aim of this tutorial is to provide the attendees with tools to evaluate and better understand active 3D sensing techniques in their own applications. Those applications may include building or purchasing a 3D system and understanding the nature of the 3D images in order to apply optimal geometric processing algorithms. This tutorial should help the attendee in modeling a given sensor into a "black box", and in understanding some image artifact that can be encountered. The case studies will expose the attendee to have a feel of some of the most interesting applications of active 3D sensing.

Intended audience

 People who are involved in the development, evaluation and application of 3D sensing techniques in an industrial or educational context.

 J.Angelo Beraldin joined the Institute for Information Technology at the National Research Council Canada in 1986. His current research interests include sensor systems engineering and analog and digital signal processing for 3-D vision systems. He has more than 40 papers and publications, 1 patent, and, has licensed technology to 3 Canadian companies. He is also a registered professional engineer with the province of Ontario.

 François Blais joined the National Research Council Canada in 1984 where he has been involved in the development of the different 3D range sensor technologies at NRC. His topics of interest cover various fields in digital signal and image processing, control, 3D vision systems and their applications. He has more than 60 papers and publications, and 8 patents of which 7 have been licensed to different Canadian industries. Mr. Blais is a registered professional engineer of Ontario.
 

Session #2       CAD/CAM Applications of 3D Optical Scanners
                       Paul Besl, Alias/Wavefront, Inc., USA

When 3D scanners are "accurate" to within ~0.1 mm, the digitized data points are directly useful in numerous CAD/CAM applications. However, many CAD/CAM applications are still not ready to receive this type of data owing to the amount of data, the "noise" issues, the occasional outliers, merging data sets, and other idiosyncrasies. This tutorial will discuss the "real-world" issues that occur when applying advanced scanner technology to design/modeling, inspection/verification, and part replication applications in today's existing CAD/CAM environments.

Benefits

If you are considering 3d optical scanning technology for your CAD/CAM
applications, this tutorial will provide you with information on what to expect in the way of the difficulties and the benefits of the technology compared to existing alternatives.

Intended audience

The presentation will not assume a background in any particular area, but some familiarity with CAD/CAM and coordinate measurement would be helpful as well as a technical background in engineering or the physical or computing sciences.

 Dr. Paul Besl has been working in this field since 1981. He is currently with Alias|Wavefront, Inc. and has prior experience with General Motors and SDRC. He received his Ph.D. from the University of Michigan, Ann Arbor, in 1986. During the last ten years, he has focused on various geometric problems in manufacturing, with a continuing involvement in the implementation of 3d optical measurement applications.

12:00-13:30   Lunch

13:30-17:30   Tutorial Sessions

Session #3        Knowledge Representation in Range Image Analysis
                        Pierre Boulanger, National Research Council Canada

The purpose is to present state-of-the-art techniques on processing range images in a coherent knowledge representation context and how to use this information to solve practical problems. The different methods used for representing surfaces and curves and how one can reliably infer them from real range images using a Baysian framework will be reviewed. From these estimated geometrical primitives, a description on how one can represent the relationships between them using various techniques and how these methods can be used will be provided. Dr. Boulanger will also demonstrate that as in complex mechanical problems in physics where a well-adapted coordinate system can significantly simplify a problem, a good knowledge representation method can also simplify the complexity of higher level processing.
 A description of how one can represent high level knowledge in a manner compatible with lower level will be addressed. He will describe how some procedural systems calling low level or intermediate level routines can interrogate and reason about the low level knowledge. Emphasis will be placed on the importance of maximum likelihood reasoning to produce a system capable of dealing with real-world situations such as false knowledge and missing information. He will also focus on feedback expectancy methods to stabilize the analysis process. During this tutorial, experimental results and an analysis of the pros and cons of each method will be presented.

 Benefits

This course will enable the attendee to grasp the basic concepts allowing the extraction of geometric primitives from 3-D data. Intended audience

Intended audience

This tutorial is primarily intended for practitioners of computer vision who are interested in state-of-the-art range image processing and understanding. The tutorial requires some basic knowledge of computer vision, differential geometry, and statistics, but these are not prerequisite.

 Dr. Pierre Boulanger is a senior research officer at the National Research Council, specializing in the field of range image processing, geometrical modeling, and virtualized reality. He has published more than 50 papers, in journals and for conferences, on various topics relating to range image processing and its applications. Dr. Boulanger has a Ph.D. from the University of Montreal in Electrical Engineering.
 

Session #4       Digitizing and modeling the human body
                       Michael Vannier, Iowa State University, USA

This tutorial will cover the various modalities for data collection in body surface anthropometry and medicine. Specific advantages and limitations of current 3D surface digitizing technology will be delineated and demonstrated.
 Visualisation, modeling and analysis software tools available will be discussed in the context of medical applications such as orthotics, prosthetics, dentistry and orthopedics. Deformable models applied to anatomic atlases, the current status of the Visible Human project (a public domain database of high resolution, 2Kx2K, 24 bit color cryosections taken at submillimeter intervals plus CT and MRI scans), and the replication of life size anatomic models for surgical planning will also be presented.

Benefits

This tutorial will offer you a broad review of the technologies involved in noncontact 3D surface digitization of the human body for anthropometry and medical applications. The information presented provides a fundamental understanding of what tools are needed to do the job, allows you to know what is currently available (avoid pitfalls and appreciate current limitations), and to be conversant with major trends and new developments in this rapidly evolving field.

 Intended audience

This tutorial is intended for researchers and engineers involve in the development of new applications in medicine and body surface anthropometry. A background in the biomedical sciences, bioengineering or ergonomics would be helpful as well as basic knowledge in computer vision.

 Dr. Michael W. Vannier is presently professor and chairman of the department of radiology at the University of Iowa. He graduated from the University of Kentucky School of Medicine and completed a diagnostic radiology residency at the Mallinckrodt Institute of Radiology at Washington University in St. Louis, Missouri. He has worked extensively on methods and applications of 3D biomedical visualization and analysis in craniofacial deformities and lower limb prosthetics. Dr. Vannier serves as editor-in-chief of the IEEE
Transactions on Medical Imaging.

18:00 - 22:00 Exhibition set - up
 

Tuesday, October 5

07:30 - 17:30  Registration

08:15 - 08:30  Conference opens  -  Welcoming Remarks

08:30 - 09:10      The Digital Michelangelo Project
                         Marc Levoy, Stanford, CA, USA, (Invited Speaker)

Session #1 - 3D Sensors (Session Chair: Dinesh Pai, UBC, Canada)

09:10 - 09:35      Hand - held acquisition of 3D models with a video camera
                         Marc Pollefeys, Reinhard Koch, Maarten Vergauwen, and Luc Van Gool, K.U.Leuven, Belgium

09:35 - 10:00      A Low - Cost Range Finder using a Visually Located, Structured Light Source
                         R. B. Fisher, A. P. Ashbrook, C. Robertson, and N. Werghi,
                         Division of Informatics, Edinburgh University, U.K.

10:00 - 10:25      Digital 3-D Imaging System for Rapid Response on Remote Sites
                        J - A Beraldin, F. Blais, L. Cournoyer, M. Rioux, S.H. El - Hakim ,
                        Visual Information Technology Group, National Research Council, Canada

10:00 - 17:00     Exhibition

10:25 - 10:55     Coffee Break

Session #2 - 3D Sensors (Session Chair: Angelo Beraldin, NRC, Canada)
10:55 - 11:20     Advances in the Cooperation of Shape from Shading and Stereo Vision
                        Holger Lange, Computing Devices, Canada
11:20 - 11:45     Moving Objects Detection from Time - Varied Background: An Application of
                        Camera 3D Motion Analysis, Zhencheng Hu and Keiichi Uchimura,
                        Computer Science Department, Kumamoto University, Japan

11:45 - 13:20  Lunch

Session #3 - 3D Sensors (Session Chair: Marc Pollefeys, K.U.Leuven, Belgium)

13:20 - 13:45      Locking onto 3D - Structure by a Combined Vergeance - and Fusion system
                         Rolf D. Henkel, Institute for Theoretical Neurophysics, University of Bremen,
                         Germany

13:45 - 14:10      3D Shape Recovery and Registration Based on the Projection of Non Coherent
                         Structured Light
                         Roberto Rodella, and Giovanna Sansoni,
                         INFM and Dept. of Electronics for the Automation, Italy

14:10 - 14:35      Curve and Surface Models to drive 3D Reconstruction using Stereo and Shading.
                         David Roussel, Patrick Bourdot, Rachid Gherbi, LIMSI/CNRS, University of Paris XI , France

14:35 - 15:05      Coffee Break and Exhibition

Session #4 - 3D Sensors (Session Chair: Paul Besl, Alias/Wavefront Inc., USA)

15:05 - 15:30      Calibration of a Laser Stripe Profiler, Alan M. McIvor,
                         Industrial Research, Auckland, New Zealand
15:30 - 15:55      Computing consistent normals and colors from photometric data
                         Holly Rushmeier and Fausto Bernardini, IBM T.J. Watson Research Center, USA

15:55 - 16:20      Bayesian Estimation of Distance and Surface Normal with a Time - of - Flight Laser Range
                         finder
                        Jochen Lang and Dinesh K. Pai, Laboratory of Computational Intelligence,
                        Department of Computer Science, University of British Columbia, Canada

16:20 - 16:45      Model - Based Scanning Path Generation for Inspection
                        Chang Shu and Fengfeng Xi, National Research Council of Canada, Canada

Poster Session (Session chair: Pierre Boulanger, NRC)

16:45 - 17:20      Poster Presentations (2 min. presentation of each poster)

17:20 - 18:30      Reception + Poster Session

Wednesday, October 6

08:00 - 17:30     Registration

08:30 - 09:10      Virtual Environments for Critical Intervention Support: Modeling, Design and
                         Implementation Issues
                         Denis Poussart, Laval University, Canada (Invited Speaker)

Session #5 - View Planning (Session Chair: Denis Poussart, Laval University)

09:10 - 09:35      CAD - based range sensor placement for optimum 3D data acquisition
                         F. Prieto , H.T. Redarce, Laboratoire d'Automatique Industrielle INSA Lyon France
                         P. Boulanger, IIT, NRC, Canada
                         R. Lepage, Laboratore d'Imagerie, de Vision et d'Intelligence Artificielle École de
                         Technologie Supérieure de Montréal, Canada

09:35 - 10:00      Computations on a Spherical View Space for Efficient Planning of Viewpoints in
                         3 - D Object Modeling
                         Ken'ichi Morooka , Hongbin Zha, and Tsutomu Hasegawa,
                         Kyushu University, Japan

10:00 - 17:00     Exhibition

10:00 - 10:25     An Automation System for Industrial 3 - D Laser Digitizing,
                        D.G. Lamb and D.L. Baird, Hymarc Ltd., Canada, M.A. Greenspan, NRC IIT, Canada

10:25 - 10:55  Coffee Break and Exhibition

Session #6 - View Registration (Session Chair: Mike Greenspan, NRC, Canada)

10:55 - 11:20     Multiview registration for large data sets
                        Kari Pulli , Stanford University, USA

11:20 - 12:45     Robust Surface Matching for Registration
                        Elizabeth Guest, COMIR, Leeds University, UK
                        Marta Fidrich, Jozsef Attila University, Szeged, Hungary
                        Steven Kelly, COMIR, Leeds University, UK
                        Elizabeth Berry, COMIR, Leeds University, UK

11:45 - 13:20  Lunch

Session #7 - View Registration (Session Chair: Frank Ferrie, McGill University)

13:20 - 13:45      Self - Calibration of a Light Striping System by Matching Multiple 3 - D Profile Maps
                         Olli Jokinen, Helsinki University of Technology, Finland

13:45 - 14:10     Registering two overlapping range images
                        Gerhard Roth, Visual Information Technology Group, National Research Council of Canada,
                        Canada

14:10 - 14:35     Fast and Robust Registration of 3D Surfaces Using Low Curvature Patches
                        Van - Duc Nguyen, GE Research & Development
                        Victor Nzomigni, CMA Consulting
                        Charles V. Stewart, Rensselaer Polytechnic Institute, USA

14:35 - 15:05     Experimental Analysis of Harmonic Shape Images
                        Dongmei Zhang and Martial Hebert ,
                        The Robotics Institute, Carnegie Mellon University, USA

15:05 - 15:30  Coffee Break and Exhibition

Session #8 - Geometric Signal Processing (Session Chair: Robert Fisher, Edinburgh University, U.K.)

15:30 - 15:55     Estimating Pose Through Local Geometry
                        Gilbert Soucy, Francesco G. Callari, and Frank P. Ferrie, Centre for Intelligent Machines,
                        Department of Electrical Engineering, McGill University, Canada

15:55 - 16:20     Efficient and Reliable Template Set Matching for 3D Object Recognition
                        Michael Greenspan and Pierre Boulanger,
                        National Research Council of Canada, Canada

16:20 - 16:45     Joined Segmentation of Cortical Surface and Brain Volume in MRI using a
                        Homotopic Deformable Cellular Model
                        Yann Cointepas, ENST  -  CNRS URA 820, France
                        Isabelle Bloch, ENST  -  CNRS URA 820 , France
                        Line Garnero, LENA  -  CNRS URA 654, France

18:30 - 21:00  Dinner  -  National Gallery of Canada

Thursday, October 7

08:00 - 17:00     Registration

08:30 - 09:10     Recent Development in Close - Range Photogrammetry
                        Armin Gruen, ETH, Zurich, Switzerland (Invited Speaker)

Session #9 - Geometric Signal Processing (Session Chair: Armin Gruen, ETH, Switzerland)

09:10 - 09:35     Extraction and tracking of surfaces in range image sequences
                        Xiaoyi Jiang, Dept. of Computer Science, Univ. of Bern, Switzerland
                        Sacha Hofer, Dept. of Computer Science, Univ. of Bern, Switzerland
                        Thomas Stahs, DaimlerChrysler AG, Research and Technology Center, Ulm, Germany
                        Ingo Ahrns, DaimlerChrysler AG, Research and Technology Center, Ulm, Germany
                        Horst Bunke, Dept. of Computer Science, Univ. of Bern, Switzerland

09:35 - 10:00      A geometric approach to the segmentation of range images
                         M.E. Bock, Dept. of Statistics, Perdue Univ. USA
                         C. Guerra, Dip. Elettronica e Informatica, Univ. Padova, Italy

10:00 - 10:25      OSCAR: Object Segmentation using Correspondence and Relaxation
                         Ben Galvin, Brendan McCane, and Kevin Novins,
                         University of Otago, New Zealand

10:25 - 10:55      Coffee Break and Exhibition

Session #10 - Object Modeling (Session Chair: Denis Laurendeau, Laval University, Canada)

10:55 - 11:20     Faithful recovering of quadric surfaces from 3D range data
                        Naoufel Werghi, Robert Fisher, Craig Robertson, and Anthony Ashbrook,
                        University of Edinburgh, U.K.

11:20 - 11:45     Progressive Multilevel Meshes From Octree Particles
                        Y. Yemez, F. Schmitt , Signal and Image Processing Department, ENST - CNRS URA820,
                        Paris -  FRANCE

11:45 - 13:20  Lunch

Session #11 - Object Modeling (Session Chair: Denis Laurendeau, Laval University, Canada)

13:20 - 13:45     Generating Smooth Surfaces with Bicubic Splines over Triangular Meshes:
                        Toward Automatic Model Building from Unorganized 3D Points
                        Toshio Ueshiba, Electrotechnical Laboratory, Japan
                        Gerhard Roth, National Research Council of Canada, Canada

13:45 - 14:10     Constructing NURBS Surface Model from Scattered and Unorganized Range Data
                        In Kyu Park, School of Electrical Engineering, Seoul National Univ. Korea
                        Il Dong Yun, School of Electronics and Control Engineering, Hankuk Univ. of F.S , Korea
                        Sang Uk Lee, School of Electrical Engineering, Seoul National Univ. Korea

14:10 - 14:35     Dynamic Gaze - Controlled Levels of Detail of Polygonal Objects in 3 - D
                        Environment Modeling
                        Hongbin Zha, Yoshinobu Makimoto, Tsutomu Hasegawa,
                        Department of Intelligent Systems, Kyushu University, Japan

14:35 - 15:05     Coffee Break and Exhibition

Session #12 - Environment Modeling (Session Chair: Martial Hebert, CMU, USA)

15:05 - 15:30     The Mapping of Texture on VR Polygonal Models
                        Denis Laurendeau, Computer Vision and Systems Laboratory
                        Department of Electrical Engineering and Computer Engineering, Laval University, Canada
                        Nathalie Bertrand and Régis Houde, Robotics Division, Hydro Québec
                        Research Institute, Canada

15:30 - 15:55      Modeling structured environments by a single moving camera
                         Repo Tapio, R"ning Juha, Infotech Oulu and Department of Electrical Engineering,
                         University of Oulu, Finland

15:55 - 16:20      Indoor Scene Reconstruction from Sets of Noisy Range Images
                        Ross T. Whitaker, Department of Electrical Engineering, University of Tennessee, USA
                        Jens Gregor and Philip F. Chen, Department of Computer Science, University of
                        Tennessee, USA

16:20 - 16:45     Large Data Sets and Confusing Scenes in 3 - D surface Matching and Recognition
                        Owen Carmichael, Daniel Huber, and Martial Hebert,
                        The Robotics Institute, Carnegie Mellon University, USA

16:45 - 17:10      Edge - based Approach to Mesh Simplification
                         Kyowoong Choo, Seoul Nat'l Univ., Seoul, Korea
                         Il Dong Yun, Hankuk Univ. of F. S., Yongin, Korea
                         Sang Uk Lee, Seoul Nat'l Univ., Seoul, Korea

Friday, October 8

08:00 - 13:00     Registration

08:30 - 09:10     The Caesar Project: A 3D Surface Anthropometry Survey
                         Kathleen Robinette, Wright - Patterson Air Force Base, Dayton, USA (Invited Speaker)

Session #13 - Human Modeling (Session Chair: Kathleen Robinette, Wright-Patterson Air Force Base, USA)

09:10 - 09:35     Building Symbolic Information for 3D Human Body Modeling from Range Data
                        L Dekker, I Douros, BF Buxton, P Treleaven,
                        Department of Computer Science, University College, London, UK
9:35 - 10:00       Building 3D Facial Models and Detecting Face Pose in 3D Space
                        LAO Shihong, OMRON Corporation, Japan
                        Yasushi SUMI, Electrotechnical Laboratory, Japan
                        Masato KAWADE, OMRON Corporation, Japan
                        Fumiaki TOMITA, Electrotechnical Laboratory, Japan
10:00 - 10:25     Automatic Body Measurement for Mass Customization of Garments
                        Andrew Certain, Manifold Graphics, Inc.
                        Werner Stuetzle, Department of Statistics, University of Washington, USA

10:25 - 10:55     Coffee Break

10:55 - 11:45     Plenary Session on Future Trends in Range Sensors and its Applications
                        Pannel members:    Paul Besl, Alias/Wavefront Inc., USA
                                                      Angelo Beraldin, NRC, Canada
                                                      Frank Ferrie, McGill University, Canada
                                                      Robert Fisher, Edinburgh University, U.K.
                                                      Martial Hebert, CMU, USA

11:45 - 13:20     Lunch

13:20                Bus departure to NRC Labs

14:00 - 16:00     Visit of NRC Visual Information Technology Laboratories
16:00                Bus departure from NRC Labs to Château Laurier Hotel.
 

List of Poster Presentations

1      3D statistical shape models for medical image segmentation
       Cristian Lorenz, Philips Research Hamburg,
       Nils Krahnstaver, Department of Computer Science and Engineering,
       Pennsylvania State University, USA

2     Hand Posture Estimation From 2D Monocular Image
       Hai-Ying Guan, Chin-Seng Chua, and Yeong-Khing Ho,
       Nanyang Technological University, China

3      On-Line Hand-Eye Calibration
       Nicolas Andreff, Bernard Espiau, and RDA Horaud,
       INRIA Rhone-Alpes/GRAVIR-IMAG, France

4      Virtual Environment Modeling by Integrated Optical and Acoustic Sensing
        V. Murino, Dipartimento Scientifico e Tecnologico, University of Verona, Italy
        A. Fusiello, R. Giannitrapani, and V. Isaia - Dipartimento di Matematica e Informatica, University of
        Udine, Italy

5      Is Appearance-Based Structure from Motion Viable?
       David E. DiFranco, Massachusetts Institute of Technology, USA
       Sing Bing Kang, Microsoft Corporation, USA

6      On Estimating the Position of Fragments on Rotational Symmetric Pottery
        Robert Sablatnig and Christian Menard, Vienna University of Technology, Pattern
        Recognition and Image Processing Group, Austria

7      Integration of Multiple Local Surface Models from Sparse Range Data
       Denis Laurendeau, Louis Borgeat, Denis Dion Jr , and Francois Cayouette,
       Computer Vision and Systems Lab, Laval University, Canada

8      Automatic Feature Correspondence for Scene Reconstruction from Multiple Views
        Philip W. Smith, Department of Electrical Engineering, University of Tennessee
        Mark D. Elstrom, Harris Corporation, Melbourne, FL, USA

9     PALM: Portable Sensor-Augmented Vision System for Large-Scene Modeling
        Teck Khim Ng and Takeo Kanade, Carnegie Mellon University, USA

10     Density and Accuracy Improvement of Phase-Based Disparity
        M. H. Ouali and D. Ziou,
        Dept. of mathematics and computer science. U. of Sherbrooke. Canada
        C. Laurgeau, Centre de Robotique. École des Mines de Paris. France

11     On the Detection of Feature Points of 3D Facial Image and Its Application to 3D Facial Caricature
        Takayuki Fujiwara, Takeshi Nishihara, Masafumi Tominaga,
        SCCS, Chukyo University, Japan
        Kunihito Kato, Faculty of Engineering, Gifu University, Japan
        Kazuhito Murakami, Faculty of of I.S.T., Aichi Prefectural University, Japan
        Hiroyasu Koshimizu, SCCS, Chukyo University, Japan

12    3D modelling and rendering of the human skeletal trunk
        S. Delorme, Y. Petit, Hôpital Sainte-Justine, Canada
        Y. Petit , Hôpital Sainte-Justine, École Polytechnique, Canada
        J.A. de Guise, École de technologie supérieure, Canada
        C.-É. Aubin, École Polytechnique, Canada
        H. Labelle, Hôpital Sainte-Justine, Canada
        C. Landry and J. Dansereau, École Polytechnique, École de technologie supérieure, Canada

13     A 3D Scanning System based on Low-Occlusion Approach
        Bor-Tow Chen, Wen-Shiou Lou Chen, Chia-Chen Chen, and Hsien-Chang Lin
        OES/ITRI, TAIWAN

14     Appearance-based virtual view generation of temporally-varying events from multi-camera images in
        the 3D Room
        Hideo Saito, Shigeyuki Baba, Makoto Kimura, Sundar Vedula, and Takeo Kanade
        Robotics Institute, Carnegie Mellon University, USA

15     Recovery of Shape and Surface Reflectance of Specular Object from Rotation of Light Source
        Hideo Saito, Kazuko Omata, and Shinji Ozawa, Keio University, Japan

16     Curvature Estimation for Segmentation of Triangulated Surfaces
        R. Sacchi, J.F. Poliakoff, The nottingham Trent University, England
        K.H. Haefele, Institut Fur Angewandte Informatik, Karlsruhe, Germany
        P.D.Thomas, The nottingham Trent University, England