🚀 go-pugleaf

  [Dr. David Kahaner is a numerical analyst on sabbatical to the 
   Office of Naval Research-Asia (ONR Asia) in Tokyo from NIST.  The 
   following is the professional opinion of David Kahaner and in no 
   way has the blessing of the US Government or any agency of it.  All 
   information is dated and of limited life time.  This disclaimer should 
   be noted on ANY attribution.]

  [Copies of previous reports written by Kahaner can be obtained using
   anonymous FTP from host cs.arizona.edu, directory japan/kahaner.reports.]

To: Distribution
From: 
 David K. Kahaner
 US Office of Naval Research Asia
 (From outside US):  23-17, 7-chome, Roppongi, Minato-ku, Tokyo 106 Japan
 (From within  US):  Unit 45002, APO AP 96337-0007
  Tel: +81 3 3401-8924, Fax: +81 3 3403-9670
  Email: kahaner@cs.titech.ac.jp
Re: Machine Vision Application (MVA'92) 7-9 Dec. 1992, Tokyo
18 Dec 1992
This file is named "mva.92"

ABSTRACT. Summary of International Association for Pattern Recognition's
conference, Machine Vision Applications (MVA'92), held in Tokyo, 7-9
Dec, 1992. 


This report was provided by

           Dr. John M. Canning
           Naito Group
           NTT Basic Research Laboratories
           3-9-11 Midori-cho
           Musashino-shi, Tokyo 180  JAPAN
            Tel: +81 422 59-4640; Fax: +81 422 59-2178
            Email: canning@apollo3.ntt.jp

Dr. Canning is working as a post-doctoral fellow in NTT's Basic Research
Laboratories.

MVA'92 is the third bi-annual conference in Japan specifically devoted
to machine vision applications.  Although it is organized by Japanese
researchers in industry and academia (along with three non Japanese
researchers), it is an international conference sponsored by the
International Association for Pattern Recognition (IAPR), and all the
presentations are in English.  The previous two conferences were also in
Japan (the first had a different name, IAPR CV'88, but was organized by
the same group).  Proceedings were printed for all three; MVA'92's is
649 pages.

>From about 240 submitted articles, the program committee chose 40 oral
and 109 poster presentations from 21 countries.  The four invited
lectures were given by Western researchers; Prof. Alex P. Pentland
(M.I.T.), Prof. Kendall Preston, Jr. (Kensall Consulting), Prof. Claus
E. Liedtke (Univ. of Hannover), and Dr. Karl Tombre (I.N.R.I.A.).  Prof.
Pentland talked about his work with modal vector shape representation
and the use of eigenvector representations of visual patterns including
human faces.  Prof. Preston discussed hardware architectures for machine
vision and voiced the opinion that future machine vision systems will
implement the sensor and processing hardware on a single chip.  Prof.
Liedtke gave an overview of machine vision applications in Germany with
a special emphasis on what types of capabilities industries were asking
for but not yet getting.

A few specific application areas drew special attention (both in special
topic sessions and the general poster sessions): map and document
understanding, face recognition and feature extraction, motion analysis
and navigation, hardware architectures for vision processors, and
printed circuit board inspection systems.  The practical capabilities in
all these areas are making good advances.  For example, Hitachi's
printed circuit board (PCB) inspection system was reported to achieve a
99.9% success rate on detecting common PCB board problems with a 0.2%
false alarm rate (although the paper cites figures of 100% and 0.06% for
one particular experiment). (First author: Takashi Hiroi, Production
Engineering Research Laboratory, Hitachi Ltd., 292 Yoshida-cho,
Totsuka-ku, Yokohama 244, Japan)

Aside from the invited presentations, I found several other
presentations of direct interest to me.  Two presentations on detecting
hand positions from image especially drew my attention.

Fukumoto, Mase, and Suenaga of NTT's Human Interface Labs have a two
camera system for following certain hand gestures of a seated operator.
The operator can point a finger to move a cursor on a large screen,
raise and lower his/her thumb to act like the button on a tabletop
mouse, and raise groups of fingers to select one of five menu choices.
(Contact: Masaaki Fukumoto, NTT Human Interface Laboratories, 1-2356
Take, Yokosuka-shi, Kanagawa-ken 238-03, Japan, 
Email: fukumoto@nttcvg.ntt.jp)

Cipolla, Okamoto and Kuno at Toshiba's Research and Development Center
have a system for tracking the motion of 4 non-coplanar points and
recovering the rigid motion of the ensemble (but not absolute
orientation and position).  In the application, the user wears a glove
or holds a wireframe model with 4 differently colored balls that can be
tracked easily by a monocular color camera.  By assuming that small
motions can be approximated by an affine transformation, three of the
points can be used to find a transformation between consecutive frames.
The difference between the actual motion of the fourth point and the
motion predicted by the affine transformation gives the motion parallax
of the fourth point.  From the motion parallax, several (but not all)
motion parameters can be computed which are enough to use the glove or
the wireframe model as a 6 dimensional mouse (position and orientation
in 3 dimensions).  One of the nicest parts about this method is that the
spatial relationships between the 4 tracked points need not be known;
the only thing that matters is that they be non-coplanar.  (Contact:
Roberto Cipolla, Dept. of Engineering, University of Cambridge,
Trumpington Street, Cambridge, CB2 1PZ England, Email: cipolla@eng.cam.ac.uk)

In document understanding, Zidouri, Chinveeraphan, and Sato of the Tokyo
Institute of Technology reported on some excellent statistical metrics
for categorizing regions of the document image as text, graphs/drawings,
or pictures.  The method uses the Minimum Covering Run representation
(MCR) of a region of a binary image which is a minimal combination of
the run-length encoding in both the horizontal and vertical directions.
(Contact: AbdelMalek B. C. Zidouri, Precision and Intelligence
Laboratory, Tokyo Institute of Technology, 4259 Nagatsuta-cho,
Midori-ku, Yokohama 227, Japan, Email: malek@pi.titech.ac.jp)

Wada and Matsuyama of Okayama University reported on a novel application
of image restoration near the crease (spine) of books scanned on flatbed
scanners or photocopiers.  Using shape from shading on the textured
cylindrical surface, they were able to restore the faded and distorted
text to a readable condition.  (Contact: Toshikazu Wada, Dept. of
Information Technology, Faculty of Engineering, Okayama University,
3-1-1 Tsushima Naka, Okayama-shi, Okayama, Japan)


The MVA'92 program follows.

  Monday, December 7, 1992

  Oral Session
  Architecture
	Chair:  Masakazu Ejiri  Hitachi, Ltd.  Japan


Kendall Preston, Jr.
New Architectures for Vision
Kensal Consulting
USA

Alberto Broggi, Vincenzo D'Andrea and Francesco Gregoretti
A Low-Cost Parallel VLSI Architecture for Low-Level Vision
*Universita di Parma, **Politecnico di Torino
Italy

Anders Astr\"om and Robert Forchheimer
MAPP2200 Smart Vision Sensor. Programmability and Adaptivity
Link\"oping University
Sweden

G. Gugliotta and A. Machi'
The Versatile Image Processor V.I.P. (Hardware Design)
IFCAI
Italy



  Oral Session
  3D Object Recognition
	Chair:  Serge Castan  Universit\'e  Paul Sabatier  France


Akira Utsumi, Yasushi Yagi and Masahiko Yachida
Estimating Surface and Spatial Structure from Wire-Frame Model Using
Geometrical & Heuristical Relations
Osaka University
Japan

Thomas M. Breuel
View-Based Recognition
IDIAP
Switzerland

John Canning
Recognizing Objects with Variable Appearances Using the VAPOR Model
NTT
Japan

Alan Wui Tze Lim, Eam Khwang Teoh and Dinesh P. Mital
A Hybrid Approach towards Segmentation of Range Images
Nanyang Technological University
Singapore

Mattias Johannesson, Anders Astr\"om and Per-Erik Danielsson
An Image Sensor for Sheet-of-Light Range Imaging
Link\"oping University
Sweden



  Poster Session 1

Yoshihiro Fujita, Nobuyuki Yamashita and Shin'ichiro Okazaki
A Real-Time Vision System Using Integrated Memory Array Processor Prototype LSI
NEC
Japan

M. Pizzocaro
An Host-Target Environment for Real Time Image Processing
Blaise Pascal University
France

Mohamed Akil and Nizar Zarka
VLSI Optimal Edge Detection Chip: Canny-Deriche Filter
Laboratoire I.A.A.I.
France

S. Ablameyko, V. Bereishik, O. Frantskevitch, M. Homenko, E. Melnik and 
  N. Paramonova
Fast Raster-to-Vector Conversion of Large-Size 2d Line-Drawings in a 
  Restricted Computer Memory
Belarusian Academy of Sciences
Belarus

Stephen M. Blackburn
Extraction of Color Region Boundaries
Australian National University
Australia

Sabine Randriamasy and Andr\'e Gagalowicz
Region-Based Top-Down Segmentation Controlled by Stereo Matching
INRIA, Rocquencourt
France

Frank Hoenes and Rainer Zimmer
Separation of Textual and Non-textual Information within Mixed-Mode Documents
German Research Center for Artificial Intelligence
Germany

Thomas Bayer
Representing and Utilising Knowledge for Understanding Structured Documents
Daimler-Benz
Germany

Qin Luo, Toyohide Watanabe and Noboru Sugie
Structure Recognition of Table-Form Documents on the Basis of the Automatic 
  Acqusition of Layout Knowledge
Nagoya University
Japan

F. Lebourgeois, J. L. Henry and H. Emptoz
An OCR System for Printed Documents
Institut National des Sciences Appliqu\'ees de Lyon
France

Miyuki Kawashima, Ryuji Tokunaga and Yuzo Hirai
Contour Map Reconstruction via Multi-Module Parallel Computational Scheme
University of Tsukuba
Japan

Katsuji Imai, Kazutoshi Gouhara and Yoshiki Uchikawa
Neural Networks Applied to Recognition of CCD Camera 
  Images where a 3-Digit Number is Stamped on the Surface of Chip Resistors
*Toyota Motor Corp., **Nagoya University
Japan

Min Seok Kang, Young Kug Ham, Hong Kyu Chung,
Rae-Hong Park and Gwi-Tae Park
Recognition of Raised Characters for Rubber Tires Classification
*Sogang University, **Korea University
Korea

Yoshitomo Yaginuma and Masao Sakauchi
Image Sequence Analysis Using Color Multiplex Image
University of Tokyo
Japan

Sanbao Xu and Per-Erik Danielsson
Robust Estimation of Focus of Expansion and Depth 
   from High Confidence Optical Flow
Link\"oping University
Sweden

Xiaoping Hu and Narendra Ahuja
Long Image Sequence Motion Analysis Using Polynomial Motion Models
University of Illinois
USA

Satoshi Abe, Koji Nakamura, Mamoru Maekawa, Takanobu Endo and Nobuyuki Sugiura
Full-Passive Human Recognition from Image Sequences
*University of Tokyo, **University of Electro-Communications
Japan

Fr\'ed\'eric Elsner, Khaled Hacine, Naceur Kerkeni, Jean-Claude Angu\'e 
   and Michel Bourton
Proposition of a Human Motion Tracking Method by Temporal-Spatial 
   Segmentation in an Image Sequence
Universit\'e de Valenciennes
France

Jian-qin Liu, Nan-ning Zheng and Xu-dong Guan
A New Perceptual Approach to Noisy X-Ray Image Segmentation
Xian Jiaotong University
PRC

Manuela Sch  Dinu Scheppelmann, Uwe Engelmann and Hans-Peter Meinzer
Segmentation of T1-, T2-, and PD-Weighted MR Images
German Cancer Research Center
Germany

Yoichi Takeshita, Mutsuhiro Terauchi and Kenji Onaga
A Method for Tumor Shape Extraction from X-Ray Mammography
Hiroshima University
Japan

Shinichiro Gomi and Koichiro Deguchi
A Hierarchical Determination of Optimal Camera and Light-Source 
   Positions for Model-Based Recognition
University of Tokyo
Japan

Fr\'ed\'eric Falzon, G\'erard Giraudon and Marc Berthod
A Multiscale Analysis Model Applied to Natural Surfaces
INRIA, Sophia-Antipolis
France

Susumu Maruno, Chow Yuet Dih,
Yasuharu Shimeki, Takashi Anezaki and Yoshikazu Okahashi
Object Recognition Using Temporal Pattern Recognition 
   Networks with Adaptive Segmentation
of Quantizer Neuron Architecture (TASQA)
*Matsushita Elec. Ind. Co., Ltd., Japan, 
  **Asia Matsushita Electric (S) Pte. Ltd., Singapore

Satoru Morita, Toshio Kawashima and Yoshinao Aoki
Generating Hierarchical Aspect Graph Using Silhouettes of Curved Objects
Hokkaido University
Japan

Daijin Kim and Minsoo Suk
Invariant Object Recognition Using Neural Network Ensemble on the CM
*Dong-A University, Korea, **Syracuse University
USA

Jun Ogata, Mikiya Sase and Yukio Kosugi
Neural Network Approaches for Attractive Area Extraction from Video Images
*Toshiba Corp., **Tokyo Institute of Technology
Japan

Masashi Koga, Tatsuya Murakami, Yoshihiro Shima and Hiromichi Fujisawa
An Extraction Method of Search Indexes for Graph Image Retrieval
Hitachi, Ltd.
Japan

Marc Viala, Christian Faye, Jean-Pierre Guerin and Didier Juvin
Cylindrical Object Reconstruction with a Moving Camera Embedded in a
``Poor'' robotic platform
*LETI, **ENSEA
France

Chao Zhang and Baozong Yuan
Representation and Reconstruction of Three-Dimensional Objects Using 
  Nonlinear Deformable Superquadric Models
Northern Jiaotong University
PR China

Tetsuo Kiuchi and Katsushi Ikeuchi
Determining Surface Roughness and Shape of Specular Diffuse Lobe 
  Objects Using Photometric Sampling Device
Carnegie Mellon University
USA

Makoto Maruya, Keiji Nemoto and Yosuke Takashima
A Texture Based Shape Connection Method for 3-D Shape Reconstruction
NEC
Japan

Makoto Kato
Theoretical Analysis of Scanning Electron Microscopes with Plural 
  Detectors as an Application Field of Photometric Stereo
Hitachi, Ltd.
Japan



  Oral Session
  Motion & Navigation
	Chair:  Robert Forchheimer  Link\"oping University  Sweden


Lampros Tsimas and Feiyu Liu
Effective Vision Algorithms for Detection of Structured and Unstructured Roads
Universit\"at der Bundeswehr M\"unchen
Germany

Michel Buffa, Olivier D. Faugeras and Zhengyou Zhang
A Complete Navigation System for a Mobile Robot, 
  Using Real-Time Stereovision and the Delaunay Triangulation
INRIA, Sophia-Antipolis
France

Xavier Leb\`egue and J. K. Aggarwal
A Mobile Robot for Visual Measurements in Architectural Applications
University of Texas at Austin
USA

Kenji Mase, Atsushi Sato, Yasuhito Suenaga and Kenichiro Ishii
A Fast Object Flow Estimation Method Based on Spacetime Image Analysis
NTT
Japan

Kenji Nagao, Masaki Sohma, Katsura Kawakami,
Michihiro Kobayakawa and Shigeru Ando
Finding Spatio-Temporal Contour of Moving Objects
*Matsushita Industrial Co., Ltd., **University of Tokyo
Japan

Alberto Del Bimbo, Paolo Nesi and Jorge L. C. Sanz
Estimation and Interpretation of Optical Flow Fields for Counting 
  Moving Objects
*University of Florence, Italy, **IBM, USA, ***IBM
Argentina

  Tuesday, December 8, 1992

  Oral Session
  Image Processing & Algorithm
	Chair:  Takashi Matsuyama  Okayama University  Japan


Katsunori Inoue and Wonchan Seo
Background Image Generation by Cooperative Parallel 
  Processing Under Severe Outdoor Condition
Osaka Univ.
Japan

Yasuko Takahashi, Akio Shio and Kenichiro Ishii
Character Extraction from Gray Images Based on Mathematical Morphology
NTT
Japan

J. E. Fischer, D. Homeister, J. Lehmler and G. Roos
High Speed Autofocus for Microscopic Images
*University of Stuttgart, **Institute for Microelectronics Stuttgart
Germany

Akira Inoue and Johji Tajima
Adaptive Image Sharpening Method Using Edge Sharpness
NEC
Japan

Jun Yamane and Masao Sakauchi
An Experiment of Realization of Fully Automated Keyword Extraction 
  in Image Database System
University of Tokyo
Japan

Pascal Legrand and Jean Pierre D\'etruin
Contour Based Image Segmentation Process on a Parallel Vision Machine
Blaise Pascal University
France



  Poster Session 2

M. Berthod, G. Giraudon and J. P. Stromboni
Massively Parallel Image Segmentation on the Connection Machine
*INRIA, Sophia-Antipolis, **Universit\'e de Nice Sophia-Antipolis
France

A. Amorouayeche
Morphological Algorithms Adapted to a Linear Acquisition
Universit\'e de Valenciennes et du Hainaut Cambr\'esis
France

Miguel A. Fern\'andez and J. Mira
Permanence Memory: A System for Real Time Motion Analysis in Image Sequences
*Universidad de Castilla la Mancha, **Faculted de Ciencias F\'isicas
de la U.N.E.D.
Spain

G. Gerardi and G. Parodi
The Programmable and Configurable Low Level Vision Unit of the HERMIA Machine
*Univ. Palerno, **Univ. Genova
Italy

Malek Adjouadi, Habibie Sumargo, Jean Andrian and Frank Candocia
The Domain of Orthogonal Transforms and the Understanding of Image Features
Florida International University
USA

Luciano da Fontoura Costa
Towards a Versatile Framework for Intermediate-Level Computer Vision
Univ. de S. Paulo
Brazil

Bruno Vasselle and G\'erard Giraudon
A Multi-Scale Regularity Measure as Geometric Criterion for Image Segmentation
INRIA, Sophia-Antipolis
France

Xu-dong Guan, Jian-qin Liu and Nan-ning Zheng
A New Parallel Scheme for Robust Segmentation of Textured Images
Xian Jiaotong University
PRC

Dinu Scheppelmann, Jochen Frey, Manuela Sch\"afer and Hans-Peter Meinzer
A Parallel Topological Map for Image Segmentation
German Cancer Research Center
Germany

J. Michael Carstensen, Michael Grunkin and Knut Conradsen
Measurement of Enzymatic Treatment Effect on Textile Using Digital 
  Image Analysis
Technical University of Denmark
Denmark

Andrew Laine and Jian Fan
Texture Classification by Wavelet Packet Signatures
University of Florida
USA

Young Kug Ham, Hong Kyu Chung, In Kwon Kim, Rae-Hong Park, Chang Bum Lee, 
  Sang Joong Kim and Byeong Nam Yoon
Hierarchical Recognition of Mixed Documents Consisting of the Korean /
Alphanumeric Texts and Graphic Images
*Sogang University, **Electronics and Telecommunications Research Institute
Korea

Rainer Hoch
Hybrid Structured Dictionary for Improving Text Recognition
German Research Center for Artificial Intelligence
Germany

Atsuhiro Takasu, Shin'ichi Satoh and Eishi Katsura
A Syntactical Approach to the Database Construction Method from Document Images
National Center for Science Information Systems
Japan

J. M. Gloger, B. Pasternak, R. Sprengel, G. Gabrielides, N. Luth 
  and M. Timmermann
An AI-Aided System for the Conversion of Paper-Based Technical 
  Drawings into a CAD Format
*Daimler Benz, **University of Hamburg, ***Institute for Production
Systems and Design Technology
Germany

Katsumi Marukawa, Masashi Koga, Yoshihiro Shima and Hiromichi Fujisawa
Error Correction for Recognition of Handwritten Kanji Names Using a 
  Name Knowledge Base
Hitachi, Ltd.
Japan

Bin-Yih Liao and Ming-Shing Young
Design of a Fuzzy Inspection and Recognition System for Binary Image
*National Kaohsiung Institute of Technology, **National Cheng-Kung University
ROC

Weiming Yao, Tsutomu Horikoshi, Takayuki Yasuno and Satoshi Suzuki
Structure from Motion Using Coarse to Fine 3d Voting
NTT
Japan

Zhen Hong and Narendra Ahuja
Target Tracking from Binocular Image Sequence Using the Autoregressive 
  Moving Average Model
University of Illinois
USA

Shuichi Nishio and Yuichi Ohta
Tracking of Vehicles at an Intersection by Integration of Multiple 
  Image Sensors
*NTT Data Communications Systems Corp., **University of Tsukuba
Japan

C. Collet, A. Quinquis and J. M. Boucher
Cloudy Sky Velocity Map Based on Matched Filter
*Ecole Navale, **ENSTBr
France

Frederic Jurie, Patrick Rives, Jean Gallice
and Jean-Luc Brame
A Vision Based Control Approach to High Speed Automatic Vehicle Guidance
*Blaise Pascal University, **INRIA, Sophia-Antipolis, ***P.S.A.
Peugeot Citro\"en
France

Haibo Li and Robert Forchheimer
Depth Independent Facial Movement Estimation
Link\"oping University
Sweden

Hirohiko Sagawa, Hiroshi Sakou and Masahiro Abe
Sign Language Translation System Using Continuous DP Matching
Hitachi, Ltd.
Japan

Hideo Tamamoto, Yuichi Narita, Akira Yanase, Futoshi Saito and Kazuto Komatsu
A Measuring System for Traffic Flow of Passers-by by Processing ITV
Image in Real Time  *Akita University, **Odate Technical High School,
***Goyo Electronics Co., Ltd., ****Ricoh System Development Co., Ltd.
Japan

Yoshitomo Yaginuma and Masao Sakauchi
A Proposal of a Multimedia Cooperative Drama Scene Recognition System
University of Tokyo
Japan

Masahisa Kaneta, Hitoshi Kanoh, Kimiharu Kanemaru and Toshio Nagai
Image Processing Method for Intruder Detection around Power Line Towers
*Hitachi Cable, Ltd., **Kansai Electric Power Co., Inc.
Japan

B. Collin and B. Zavidovique
Under the Sun Light. From Physics to Image Processing
*ETCA/CREA/SP, **Univ. de Paris-Sud Orsay
France

Masataka Ajiro, Hiroyuki Miyata,
Takashi Kan and Makoto Ono
Satellite Image Processing Using Cellular Array Processor(CAP)
*ERSDAC Data Center, **Mitsubishi Electric Corporation
Japan

Daisuke Yagi, Keiichi Abe and Hiromasa Nakatani
Segmentation of Color Aerial Photographs Using HSV Color Models
Shizuoka University
Japan

Katsunori Inoue, Shuichi Fukuda, Masashi Okubo and Tong Qin
Model Based Object Recognition Using Modified Coded Boundary 
  Representation (MCBR) Method
*Osaka University, **Tokyo Metropolitan Institute of Technology, Japan,
   ***Harbin Institute of Technology
PRC

Seiji Ishikawa, Keiichi Sato, Predrag Minovic and Kiyoshi Kato
An Interactive 3d Symmetry Analysis System
*Kyushu Institute of Technology, **SECOM Co. Ltd.
Japan

Yasuo Seki and Hisayuki Tatsumi
Stereo Matching Based on Iterative Incompatibility Revision
Kanagawa Institute of Technology
Japan

Bikash Sabata and J. K. Aggarwal
Correspondence of Surfaces in a Sequence of Range Images 
  for Motion Estimation and Tracking
University of Texas at Austin
USA

Takio Kurita and Pierre Boulanger
Computation of Surface Curvature from Range Images Using Geometrically 
  Intrinsic Weights
*Electrotechnical Laboratory, Japan, **National Research Council Canada
Canada



  Oral Session
  Map & Drawing
	Chair:  Narendra Ahuja  University of Illinois  USA


Karl Tombre
Technical Drawing Recognition and Understanding: From Pixels to Semantics
INRIA, Lorraine
France

Peter Mowforth
Data Conversion for GIS
The Turing Institute Ltd.
UK

Shin'ichi Satoh and Masao Sakauchi
Understanding Rule Generation Supporting System for Drawing 
  Understanding using Interaction with User
*National Center for Science Information Systems, **University of Tokyo
Japan

Gerhard Roth
Primitive Extraction Using Incremental Curve Generation
National Research Council of Canada
Canada



  Oral Session
  Applications 1
	Chair:  Claus-E. Liedtke  Universit\"at Hannover  Germany


Curt L. Orbert, Ewert W. Bengtsson and Bo G. Nordin
Practical Considerations about a 2d Algorithm for Object Orientation
in Industrial Applications
Uppsala University
Sweden

Hideaki Doi, Yasuhiko Hara, Koichi Karasaki,
Tadashi Iida, Takashi Furutani, Shigeki Kitamura,
Norihiro Minatani and Satashi Shinada
Automated Inspection of Printed Circuit Board Patterns Referenced to CAD Data
*Hitachi, Ltd., **Hitachi Video and Information System, Inc.
Japan

Saburo Okada, Masaaki Imade, Hidekazu Miyauchi, Tetsuhiro Sumimoto 
  and Hideki Yamamoto
Detection and Discrimination of Surface Defects by Analyzing
Diffraction Pattern of Laser Beam
*Government Industrial Research Institute, Chugoku, 
  **Maritime Safety Academy, ***Okayama University
Japan

Takashi Hiroi, Kazushi Yoshimura, Takanori Ninomiya, Toshimitsu Hamada, 
  Yasuo Nakagawa, Shigeki Mio, Kouichi Karasaki and Hideaki Sasaki
Development of Solder Joint Inspection Method Using Air Stimulation
Speckle Vibration Detection Method and Fluorescence Detection Method
Hitachi, Ltd.
Japan



  Wednesday, December 9, 1992

  Oral Session
  3D Shape Analysis & Recognition
	Chair:  Kendall Preston, Jr.  Kensal Consulting  USA


Alex P. Pentland
Modal Descriptions for Recognition and Tracking
Massachusetts Institute of Technology
USA

Kenichi Kanatani
Statistical Reliability of 3-D Computing from Images
Gunma University
Japan

Latifa Guisser, Ren\'e Payrissat and Serge Castan
An Accurate 3d Vision System Using a Projected 
  Grid for Surface Geometrical Study
Universit\'e Paul Sabatier
France

Tomas Gustavsson and Quan Liang
Three Dimensional Boundary Detection Using Higher-Order 
  Surface Fitting and Directional Smoothing
Chalmers University of Technology
Sweden

Yoshinao Aoki, Shin Tanahashi, Osamu Kinoshita and Hideaki Nakamura
Development of a System for Producing Stereo Ground Models by Optical 
  Lithography
*Hokkaido University, **Hokkaido Chizu Co. Ltd.
Japan



  Oral Session
  Human Motion & Face Analysis
	Chair:  Alex Pentland  Massachusetts Institute of Technology  USA


Shizuo Sakamoto, Yoko Miyao and Johji Tajima
Extracting Feature Points on Human Eye Photographs
*NEC, **NEC Software Limited
Japan

Shigeru Akamatsu, Tsutomu Sasaki, Hideo Fukamachi and Yasuhito Suenaga
Target Image Extraction for Face Recognition Using the Sub-Space 
  Classification Method
*ATR, **NTT, ***NTT Software Corp.
Japan

Yasushi Sumi and Yuichi Ohta
Human Face Analysis Based on Distributed 2d Appearance Models
University of Tsukuba
Japan

Masaaki Fukumoto, Kenji Mase and Yasuhito Suenaga
Real-Time Detection of Pointing Actions for a Glove-Free Interface
NTT
Japan

Roberto Cipolla, Yasukazu Okamoto and Yoshinori Kuno
Qualitative Visual Interpretation of 3d Hand Gestures Using Motion Parallax
Toshiba Corp.
Japan



  Poster Session 3

Mohamed Akil and Eric Dujardin
A 'Hyper-Pyramid' Architecture for Massively Parallel Image Processing
Laboratoire I.A.A.I.
France

Hideto Motomura, Katsuhiro Kanamori, Teruo Fumoto and Hiroaki Kotera
Applications of the Color Conversion System Using the LUT and 
  Interpolation to the Real-Time Color Recognition
Matsushita Research Institute Tokyo, Inc.
Japan

Min Xue, Abdelhamid Hachicha and Alain M\'erigot
An Efficient Parallel Implementation of the Laplacian Pyramid Algorithm
*ESIGETEL, **Universit\'e Paris-Sud
France

Balaram Bhattacharyya
A Rule-Based Expert System for Low Level Image Processing
Visva-Bharati University
India

Hiroyasu Koshimizu and Munetoshi Numada
On a Basic Consideration of the Warp Model of Hough Transform
*Chukyo University, **Lossev Technology Co. Ltd.
Japan

AbdelMalek B. C. Zidouri, Supoj Chinveeraphan and Makoto Sato
A Stroke Index for Document Image Analysis Based on the MCR Expression Method
Tokyo Institute of Technology
Japan

Masaharu Ozaki and Yuusuke Ishida
A Document Analysis Method Based on a Consistent Structural 
  Model of Document Elements and Pages
*Fuji Xerox Palo Alto Lab., USA, **Fuji Xerox Co.,Ltd.
Japan

Koichi Kise, Tadamichi Shiraishi, Shinobu Takamatsu and Hiroji Kusaka
Improvement of Text Image Recognition Based on Linguistic Constraints
University of Osaka Prefecture
Japan

Wookhyun Kim, Takashi Furukawa, Yuzo Hirai
and Ryuji Tokunaga
Extraction and Reconstruction of Road Segments by Spatial Filters
*University of Tsukuba, **Nippon Steel Corporation
Japan

Hisao Niwa, Kazuhiro Kayashima and Yasuharu Shimeki
Postprocessing for Character Recognition Using Keyword Information
Matsushita Elec. Ind. Co., Ltd.
Japan

Minami Miyauchi, Masatoshi Seki,
Akira Watanabe and Arata Miyauchi
Interpretation of Optical Flow through Neural Network Learning
*SANNO College, **Musashi Institute of Technology
Japan

Satoshi Abe, Taizo Okano, Hisao Sato,
Koki Sengoku, Nobuaki Shimada and Noboru Kazuki
A High-Speed Image Processor for Detection of Pavement Cracks
*University of Tokyo, **Array Corporation
Japan

Mo Weiguo
Combining Motion Smoothness and Greyscale Consistency in 
  Analysing Real World Human Body Motion
Fudan University
PR China

Akihiko Sugikawa
A New Multi-Stage Face Identification in Motion Picture
Toshiba Corp.
Japan

Joan Batlle Grabulosa and Antonio B. Mart\'inez
Detection of Scene Changes for Efficient Image Transmission 
  from a Moving Camera
*Universitat de Girona, **Universidad Polit\`ecnica de Catalunya
Spain

Hideyuki Ban, Akihide Hashizume and Ryuuichi Suzuki
Interactive Surface Display for 3d Medical Images Using Multilayer 
  Range Images
Hitachi, Ltd.
Japan

Myl\`ene Roussel, X. W. Tu and D. Fontaine
Biological Object Classification and Identification on Light Microscopy Images
*Universit\'e de Technologie de Compi\`egne,  **Laboratoire
d'Informatique Avanc\'ee de Compi\`egne
France

Kazuhiko Fukuda, Makoto Watanabe, Masao Nitoh, Yuh Maruko and Youji Suzuki
Agricultural Products Sorting System by Using High-Speed Image Pipelined 
  Processor and FA Components
*FUJI FACOM Corporation, **FUJI Electric Co., Ltd.
Japan

Gunilla Borgefors and Gabriella Sanniti di Baja
Parallel Analysis of Non Convex Shapes Digitized on the Hexagonal Grid
*Swedish Defence Research Establishment, Sweden, **Istituto di Cibernetica
Italy

T. W. Rauber and A. S. Steiger-Gar\c c\~ao
2-D Form Descriptors Based on a Normalized Parametric Polar Transform 
  (UNL Transform)
Universidade Nova de Lisboa
Portugal

Katsuyuki Kamei, Yasuaki Nakamura and Shigeru Abe
A Real-Time Deformable Clay Model for Rotationally Symmetric Objects
Mitsubishi Electric Corporation
Japan

R. K. Kunchev and M. G. Milanova
A CAD Based Knowledge Vision System for 3d Objects Recognition
Technical University
Bulgaria

Hongbin Zha and Tadashi Nagata
3-D Curved Object Recognition Based on Parameter Nets Weighted 
  with Feature Saliency
Kyushu University
Japan

Raj Talluri and J. K. Aggarwal
Transform Clustering for Model-Image Feature Correspondence
University of Texas at Austin
USA

Morito Shiohara and Hiroshi Kamada
Illumination-Invariant Active Vision
FUJITSU Laboratories Ltd.
Japan

Yoshiharu Morimoto and Yutaka Yoshioka
3-D Shape Measurement by Extraction of Characteristics Using Image Processing
Osaka University
Japan

Toshikazu Wada and Takashi Matsuyama
Shape from Shading on Textured Cylindrical Surface --- Restoring Distorted 
  Scanner Images of Unfolded Book Surfaces ---
Okayama University
Japan

Yoshifumi Kitamura, Takeo Kimura, Masatoshi Okutomi and Hideyuki Tamura
Surface Reconstruction with Triangular Patches from Multiscale Range Images
Canon Inc.
Japan

Tetsuo Miyake, Ken Uemura, Shigeyuki Seto, Yasunari Saito and Kazuaki Shimizu
A Technique for Reconstructing Shape of Specular Surfaces
*Toyohashi University of Technology, **Asahi Glass Co. Ltd.
Japan

Jesse S. Jin, Wai-kiang Yeap and Brian G. Cox
Recovering Depth in Stereo Calculation
University of Otago
New Zealand

Hironobu Takahashi, Hideo Hirono, Hajime Terasaki and Fumiaki Tomita
Primitive Based Stereo for the Can-Picking Robots
*Sanyo Electric Company, **Electrotechnical Laboratory
Japan

E-ren Chuang and David Sher
Implementation of an Object-Oriented & Declarative Language for 
  Image Understanding
*Kaohsiung Polytechnic Institute, ROC, **State University of 
  New York at Buffalo
USA

Charles Allen and Roy Booth
A Monocular Technique for the Reconstruction of 3d Shape 
  Using a Cooperating Robot Arm
University of Newcastle upon Tyne
UK

Jianqin Liu, Nanning Zheng, Li Wei and Qingyuan Wang
A Neural Network Based Method for Automated 
  Classification of Sheep Leather Images
Xian Jiaotong University
PRC



  Oral Session
  Applications 2
	Chair:  Karl Tombre  INRIA, Lorraine  France


Claus-E. Liedtke
New Machine Vision Applications in Germany
Universit\"at Hannover
Germany

Y. Jin, Y. Goto, H. Naito, Y. Aono and A. Iwake
In-Motion Inspection of Obstacles around Railways 
  by a Highly Accurate Laser-Sectioning Method
*Kobe Steel Ltd., **Central Japan Railway Co.
Japan

Jeong-Kee Kim, Chan-Eung Park and Kwae-Hi Lee
Automatic Digital Elevation Model Extraction Using SPOT Satellite Image
Sogang University
Korea

Gopal Sarma Pingali and Ramesh Jain
Probe Shape Recovery in Scanning Probe Microscopy
University of Michigan
USA

M. Chapron, N. Ivanov, P. Boissard and P. Valery
3d Reconstruction of Plants
*ENSEA-ETIS, **INRA Bioclimatologie
France


------------------------------END OF REPORT--------------------------