ICASSP-94
`
`VOLUMES
`
`I
`
`IMAGE AND MULTIDIMENSIONAL
`SIGNAL PROCESSING
`
`1994
`IEEE INTERNATIONAL CONFERENCE
`ON
`ACOUSTICS, SPEECH AND
`SIGNAL PROCESSING
`
`APRIL 19-22, 1994
`ADELAIDE CONVENTION CENTRE
`ADELAIDE, SOUTH AUSTRALIA
`
`IEEE
`94CH3387-8
`
`Sponsored by
`THE INSTITUTE OF ELECTRICAL AND
`ELECTRONICS ENGINEERS,
`SIGNAL PROCESSING SOCIETY
`
`Page 1 of 12
`
`GOOGLE EXHIBIT 1026
`
`

`

`"J tuwH-
`-,p
`3'1'.J
`:Cto2.
`
`---/'lq -5
`
`ICASSP-94 Proceedings in 6 Volumes:
`
`Volume 1.
`
`Volume 2.
`
`s,
`
`S2AUVN
`
`Volume 3.
`
`Volume 4.
`
`Volume 5.
`
`Volume 6.
`
`D
`s
`
`I
`
`p
`
`Speech Processing 1
`
`Speech Processing 2
`Audio
`Underwater Acoustics
`VLSI
`Neural Networks
`
`Digital Signal Processing
`
`Statistical Signal and Array Processing
`
`Image and Multidimensional Signal Processing
`
`Plenary and Special Sessions
`
`Cover Design: David Jones Art and Design, Adelaide, South Australia
`Typesetting: Musictype Pty. Ltd., Adelaide, South Australia
`
`Additional copies may be ordered from:
`
`IEEE Service Center
`445 Hoes Lane
`P.O. Box 1331
`Piscataway, NJ 08855-1331
`
`Copyright and Reprint Pennission: Abstracting is pennitted with credit to
`the source. Libraries are pennitted to photocopy"beyond the limit of U.S.
`copyright law for private use of patrons those articles in this volume that
`carry a code at the bottom of the first page, provided the per-copy fee
`indicated in the code is paid through Copyright Clearance Center, 27
`Congress Street, Salem, MA 01970. For other copying, reprint or
`republication pennission, write to IEEE Copyrights Manager, IEEE
`Service Center, 445 Hoes Lane, P.O. Box 1331, Piscataway, NJ 08855-
`1331. All rights reserved. Copyright © 1994 by the Institute of Electrical
`and Electronics Engineers, Inc.
`
`IEEE Catalog No. 94CH3387-8
`ISBN 0-7803-1775-0 (softbound)
`ISBN 0-7803-1776-9 (casebound)
`ISBN 0-7803-1777-7 (microfiche)
`Library of Congress Catalog Card No. 84-645139 (serial)
`
`Page 2 of 12
`
`

`

`VOLUME 5
`
`IMAGE & MULTIDIMENSIONAL
`SIGNAL PROCESSING
`Talile of Contents
`
`IMAGE ANALYSIS I
`
`Chairperson: Ed Delp, Purdue University, (USA)
`
`Unsupervised Segmentation of Radar Images Using Wavelet
`Decomposition and Cumulants
`Jean-Marc Boucher, Stepltane Pleihers, ENSTB, (FRANCE)
`
`Morphological Scale-Space Fingerprints
`and their Use in Object Recognition in Range Images
`Paul Jackway, Wageeh Boles, Mohamed Deriche, Signal Processing
`Research Centre, Queensland University of Technology, (AUSTRALIA)
`
`Detection of Point Targets in High Resolution
`Synthetic Aperture Radar Images
`Ying Wang, Rama Chellappa, Qinfen Zheng, University of Maryland,
`(USA)
`
`V-1
`
`V-S
`
`V-9
`
`An Energy Minimization Approach to Building Detection
`in Aerial Images
`',
`Santhana Krishnamachari, Rama Chellappa, University of Maryland,
`(USA)
`
`V -13
`
`Use ofM-Band Wavelet Transform for Multidirectional
`and Multiscale Edge Detection
`Turgut Aydin, Yiicel Yemez, Billent Sankur, Emin Anarim, Oktay Aikin,
`Bogazir;i Universitesi, (TURKEY)
`
`V-17
`
`V-21
`A Wavelet Based Mammographic System
`Andrew Laine, University of Florida. Mike Lewis, The Athena Group,
`Fred J Taylor, University of Florida, (USA)
`
`Orientation Selective Operators for Ridge, Valley, Edge and
`Line Detection in Imagery
`Jianxin Hou, Robeno H Bamberger, Washington State University, (USA)
`
`V-25
`
`Periodicity Estimation in Textured Images Using a ML
`Approach
`Jorge Marques, IN ESC/Technical University of Lisbon, ( PORTUGAL)
`
`V-29
`
`Detecting Scene Changes and Activities in Video Databases
`Rohen Hsu, Hiroshi Harashima, University of Tokyo, (JAPAN)
`
`On Vector Quantization for Fast Facet Edge Detection
`MY Jaisimha, Jill R Goldschneider, Alexander E Mohr,
`Eve A Riskin, Rohen M Haralick, University of Washington, Seallle,
`(USA)
`
`A Contour-Based Part Segmentation Algorithm
`Mohammed Bennamoun, Signal Processing Research Centre,
`Queensland University of Technology, (AUSTRALIA)
`
`V-33
`
`V-37
`
`V-41
`
`An Improvement on the Reception of PAL-M Television
`Signal Using an Additional Simple Delay Filter
`Yuzo lano, University of Campinas. (BRAZIL)
`
`V-49
`
`Image Visual Quality Restoration by Cancellation of the
`Unmasked Noise
`Benoit Macq, Universitl Catholique de Louvain, (BELGIUM) ; Marco
`Mattavelli, Inst. Polyt. Fed. de lousanne, (SWITZERLAND); Olivier Van
`Caister, Emmanuel van der Piancke, Serge Comes, Universitl Catholique
`de louvain (BELGIUM); Wei Li, Inst. Polyt. Fed. de Lausanne
`(SWITZERLAND)
`
`V-S3
`
`Processing
`Filtering
`Carl-Fredrik
`
`Subspace
`
`Ji•~~~~bJ~~i:t,,~r,iij~iii
`
`V-S7
`A 2-D IIR Neural Hybrid Filter for Image Processing
`Mitsuji Muneyasu, Satoshi Tsujii, Takao Hinamoto, Hiroshima University,
`(JAPAN)
`
`Adaptive ex. Trimmed Mean Filters with Excellent
`Detail-Preserving
`Akira Taguchi, Musashi Institute ofTechnology, (JAPAN)
`
`Design of Optimal Median-Type Filters under Structural
`Constraints
`Bing Zeng, Hong Kong University of Science & Technology,
`(HONGKONG)
`
`Statistical Analysis of the Median Based Multi-Shell
`Order-Statistics Filters
`JS Jimmy Li, Anand Ramsingh, University of Canterbury,
`(NEW ZEALAND)
`
`GLMOS and a Comparative Study of Nonlinear Filters
`Hamid Rabiee, R L Kashyap, Purdue University, (USA)
`
`V-61
`
`V-65
`
`V-69
`
`V-73
`
`Statistical Morphological Filters for Binary Image Processing
`Carlo Regazzoni, University of Genova, (ITALY); Anastasios N
`Venetsamopoulos, University ofToronto, (CANADA) ; Gianluca Foresti,
`Gianni Vernazza, University of Genova, (ff ALY)
`
`V -77
`
`A Human-Machine Interactive System for Efficient
`Image Restoration
`Hong Tang, Australian National University (AUSTRALIA)
`
`V-81
`
`Blind Superresolving Image Recovery from Blur-Invariant Edges V-85
`Kazuki Nishi, University of Electro-Communications, Shigeru Ando,
`University of Tokyo, (JAPAN)
`
`IMAGE DISPLAY
`
`IMAGE FILTERING AND ENHANCEMENT
`
`Chairperson: Ping Wah Wong, Hewlett-Packard laboratories, (USA)
`
`Chairperson: P Combettes, City University of New York ( USA)
`
`An Image Processing Algorithm for a Super High Dennition
`Imaging Scheme with Multiple Different-Aperture Cameras
`Takahiro Saito, Takashi Komatsu, Kanagawa University,
`Kiyoharu Aizawa, The University ofTokyo, (JAPAN)
`
`V-45
`
`Colour Quantization of Images Based
`on Human Vision Perception
`Navin Chaddha, Wee-Chiew Tan, Teresa H Y Meng, Stanford University,
`(USA)
`
`V-89
`
`Page 3 of 12
`
`

`

`Colour Quantization Error in Terms of Perceived
`Image Quality
`Alain Tremeau, URA 842 Lab TS/, Maurice Calonnier, lnstitut Textile de
`France, Bernard Laget, URA 842 Lab TS/, (FRANCE)
`
`V-93
`
`Greedy Tree Growing for Colour Image Quantization
`Tsann-Shyong Liu, Long-Wen Chang, Tsing Hua University ,
`(TAIWAN ROC)
`
`V-'.17
`
`Designing the Colour Palette for Textile Materials
`Printing Process
`Gabriel Marcu, Kansei Iwata, Graphica Computer Corporation, (JAPAN)
`
`V-101
`
`Halftoning Technique Using Genetic Algorithm
`Naoki Kobayashi, Hideo Saito, Keio University, (JAPAN)
`
`V-105
`
`Conversion of Scanned Documents to the Open
`Document Architecture
`Gary Farrow, Costas Xydeas, John Oakley, Manchester University, (UK)
`
`V-109
`
`Error Diffusion with Dynamically Adjusted Kernel
`Ping Wah Wong, Hewie/I-Packard Laboratories, (USA)
`
`V-113
`
`A Distortion Measure for Image Artifacts
`Based on Human Visual Sensitivity
`Shanika Karunasekera, Nick Kingsbury, Cambridge University , (UK)
`
`V-117
`
`Application of Directional Statistics in Vector Direction
`Estimation
`Nikos Nikolaidis, Ioannis Pitas, University ofThessaloniki, (GREECE)
`
`V-121
`
`Generating Non-Gaussian Random Fields for Sea-Surface
`Simulations
`Garry N Newsam, Michael Wegener, Defence Science Technology
`Organisation, (AUSTRALIA)
`
`•
`
`IMAGE ANALYSIS II
`
`Chairperson: Rosalind Picard, Mrr Media Laboratory, (USA)
`
`On-Line Cursive Handwriting Recognition Using
`Speech Recognition Methods
`Thad Stamer, MIT Media Lab, John Makhoul, Richard Schwartz,
`George Chou, Bolt Beranek and Newman Inc, (USA)
`
`V-125
`
`A New Wold Ordering for Image Similarity
`Rosalind W Picard, Fang Liu, MIT Media Laboratory, (USA)
`
`V-129
`
`Filter Estimation Maximization Algorithm for
`Image Segmentation
`Hocine Cherifi, TS/ (CNRS 842), Richard Grisel, /CPI - L2S2, (FRANCE)
`
`V-133
`
`Recognition of Space Curves Based on the Dyadic
`Wavelet Transform
`Quang Minh Tieng, Wageeh Boles, Signal Processing Research Centre,
`Queensland University ofTechnology, (AUSTRALIA)
`
`V-137
`
`Modelling and Classification of Shapes in Two-Dimensions
`Using Vector Quantization
`Simon Lee, Brian Lovell, University of Queensland/CSSIP, (AUSTRALIA)
`
`V-141
`
`V-145
`Heuristic Image Decoding Using Separable Source Models
`Anthony C Kam, Caliper Corp, USA; Gary E Kopec, Xerox PARC, (USA)
`
`Supervised Hidden Markov Modelling for On-line
`Handwriting Recognition
`Jerome R Bellegarda, David Nahamoo, Krishna Nathan, Eveline
`Bellegarda, IBM T J Watson Research Centre, (USA)
`
`V-149
`
`Correlation Filters for Texture Recognition and Applications
`to Terrain-Delimitation in Wide-Area Surveillance
`Hemani Singh, Aware Inc , Abhijit Mahalanobis, Hughes Missile Systems
`Co., (USA)
`
`V-153
`
`COMPUTED IMAGING I:
`SYNTHETIC APERTURE
`
`Chairperson: Mehrdad Soumekh,
`State University of New York al Buffalo, (USA)
`
`Optimal Configuration and Weighting of Nonuniform Arrays V-157
`According to a Maximum ISLR Criterion
`Carlo Boni, Mario Richard, Alenia S.pA. , Sergio Barbarossa,
`University of Rome "La Sapienza", (ff ALY)
`
`Target Shifts Due to Modelling Assumptions In Inverse
`Synthetic Aperture Radar
`Hyeokho Choi, David C Munson Jr, University of Illinois at Urbana(cid:173)
`Champaign, (USA)
`
`V-161
`
`Comparative Study of Some Algorithms for Terrain
`Classification Using SAR Images
`Zied Belhadj, S2HF - /REST£, Ali Saad, LAT/ - IRESTE,
`Safwen El Assad, J Saillard, S2HF - IRESTE, Dominique Barba,
`LAT/ - IRESTE, (FRANCE)
`
`V-165
`
`Automatic Target Detection in Dynamic Clutter from
`Incoherent ISAR Data
`Mehrdad Soumekh, State University of New York al Buffalo,
`Michael Pollock, Robert Dinger, Naval Command, Control and Ocean
`Surveillance Center, (USA)
`
`V-169
`
`Imaging of Multitargets with ISAR Based on the
`Time-Frequency Distribution
`Aiyuan Wang, Yinfang Mao, Zongzhi Chen, Institute of Electronics
`Academia Sinica, (PR CHINA)
`
`V-173
`
`ISARLAB: A Radar Signal Processing Tool
`Brett Haywood, Anthony Zyweck, Ross Kyprianou, Defence Science
`and Technology Organisation!CSSIP, (AUSTRALIA)
`
`V-177
`
`A New Methodology for Fourier Synthesis.
`Fourier Interpolation and Reconstruction via Shannon-type
`Techniques: FIRST
`Andre Lannes, Eric Anterrieu, Sylvie ~oques, Geraldine Fitoussi,
`CNRSIOMP, (FRANCE)
`
`V-181
`
`Synthetic Aperture Technique Used in Ultrasonic Intravascular V-185
`Imaging
`Jiang Hui, Hou Chao-huan, Academia Sinica, (PR CHINA)
`
`V-189
`Vector Quantization of Raw SAR Data
`Jean-Marie Moureaux, Patricia Gauthier, l3S-CNRS, Michel Barlaud,
`Pascale Bellemain, Aerospatiale (Cannes), (FRANCE)
`
`A Real Time Processor for the Australian Synthetic
`Aperture Radar
`,
`Nick Stacy, Michael Burgess, J J Douglass, Mars~ll Muller, Murray
`Robinson, Defence Science and Technology Organisation, (AUSTRALIA)
`
`V-193
`
`Sign Language Image Processing for an Intelligent
`Communications by a Communication Satellite
`Yoshinao Aoki, Shin Tanahashi, Jun Xu, Hokkaido University, (JAPAN)
`
`V-l'.17
`
`MOTION ESTIMATION
`
`r
`Chairperson: Rama Chellappa, University of Maryland, (USA)
`\
`
`V-201
`A New Approach to Motion Estimation for ISAR Imaging
`Stephen Simmons, Robin Evans, University of Melbourne, (AUSTRALIA)
`
`Motion Compensated Video Sequence Interpolation Using
`Digital Image Warping
`Jacek Nieweglowski, Timo Moisala, Tampere University of Technology,
`Petri Haavisto, Nokia Research Center, (FINLAND)
`
`V-205
`
`Page 4 of 12
`
`

`

`V-209
`Motion Estimation Using Multiple Image Sensors
`Kiyoharu Aizawa, Ken-ichi Iwata, University of Tokyo, Takahiro Saito,
`Kanal(awa Unil'ersity, Mitsutoshi Hatori, University of Tokyo, (JAPAN)
`
`Object-Oriented Video Coding Employing Dense
`Motion Fields
`Christoph Stiller, Aachen University of Technology, (GERMANY)
`
`V-273
`
`Spanning the Gap Between Motion Estimation and Morphing V-213
`Michele Covell, Margaret Withgoll, Interval Research Corporation,
`(USA)
`
`3D Contour Image Coding Based on Morphological Filters
`and Motion Estimation
`Chuang Gu, Swiss Federal Institute o/Teclmology, (SW/7ZERLAND)
`
`V-277
`
`A New Technique for Block-Based Motion Compensation
`Shinichi Kozu, NEC Corporation, (JAPAN); Sanjeev Kulkarni,
`Princeton University, (USA)
`
`V-217
`
`An Algorithm for Simultaneous Motion Estimation
`and Scene Segmentation
`Michael M Chang, University of Rochester, M Ibrahim Sezan, fastman
`Kodak Co, A Murat Tekalp, University of Rochester, (USA)
`
`V-221
`
`An Analogue Interpretation of Compression for Digital
`Communication Systems
`John M Lervik, Tor A Ramstad, Norwegian Institute of Technology,
`(NORWAY)
`
`V-281
`
`A Fast Algorithm for Region-Oriented Texture Coding
`Marco Cermelli, Fabio Lavageuo, Maueo Pampolini,
`DIST- University of Genova, (ITALY)
`
`V-285
`
`Digital Video Standards Conversion in the Presence of
`Accelerated Motion
`Andrew J Paui, University of Rochester, M Ibrahim Sezan,
`Eastman Kodak Co, A Murat Tekalp, University of Rochester, (USA)
`
`V-22S
`
`Address Predictive Colour Quantization Image
`Compression For Multimedia Applications
`Lai-Man Po, Wen-Tao Tan, Chi-Ho Chan, City Polytechnic of Hong Kong,
`(HONGKONG)
`
`V-289
`
`Nonuniform Image Motion Estimation Using Kalman Filtering V-229
`Nader Namazi, Pablo Penafiel, Chieh-Min Fan, The Catholic
`Unfrersity of America, ( USA)_
`
`Intensity Scale Invariant Motion Estimation with Rotation and V-233
`Spatial Scaling Information
`Colin Bussiere, Dimitrios Hatzinakos,. University of Toronto, (CANADA)
`
`COMPUTED TOMOGRAPHY
`
`Chairperson: David Munson, University of Illinois, (USA)
`
`Genetic Algorithms for Neuromagnetic Source Reconstruction V-293
`Paul S Lewis, John Mosher, las Alamos National laboratory, (USA)
`
`Restoration of Low Bit Rate Compressed Images using
`Mean Field Annealing
`James C Brailean, Motorola, Taner Ozcelik, Aggelos K Katsaggelos,
`Northwestern University , (USA)
`
`V-237
`
`Image Reconstruction of Contour Data Using a
`Backpropagation Neural Network
`Karim Faez, Amirkabir University o/Technology, (/RAN);
`Mohamed Kamel, University of Waterloo, (CANADA)
`
`V-297
`
`Adaptive Multi-Feature Motion Estimation
`Regis Crinon, Wojciech Kolodziej, Oregon State University, (USA)
`
`V-241
`
`Replacement Noise in Image Sequences -
`Detection and Interpolation by Motion Field Segmentation
`Robin Morris, William Fitzgerald, Cambridge University, (UK)
`
`V-24S
`
`IMAGE CODING O AND QUANTIZATION
`
`Chairperson: Tor Ramstad,
`Norwegian Institute o/Teclmology, (NORWAY)
`
`Minimum Generalised Quadratic Error Quantization
`for Image and Video Coding
`John Princen, AT & T Bell laboratories, (USA); Ming H Chan, Telecom
`Australia, (AUSTRALJA)
`
`V-249
`
`Entropy-Constrained Predictive Trellis Coded Quantization:
`Application to Hyperspectral Image Compression
`Glen P Abousleman, Michael W Marcellin, Bobby R Hunt,
`University of Arizona, (USA)
`
`V-253
`
`Image Coding Using Adaptive Recursive Interpolative DPCM V-257
`with Entropy-Constrained Trellis Coded Quantization
`Eric Gifford, Bobby R Hunt, Michael Marcellin, University of Arizona,
`(USA)
`
`Deterministic Annealing for Trellis Quantizer
`and HMM Design Using Baum-Welch Re-Estimation
`David Miller, Kenneth Rose, University of California at Santa Barbara,
`Philip A Chou, Xerox Palo Alto Research Center, (USA)
`
`V-261
`
`Sampling of Two-Dimensional Signals Below Nyquist Density
`with Applications to Computer Aided Tomography
`Kai-Kou Roger Yu, Sze-Fong Yau, Hong Kong University of Science
`and Technology, (HONG KONG)
`
`V-301
`
`A Fast Tomographic Reconstruction Algorithm in the 2-D
`Wavelet Transform Domain
`Laure Blanc-Feraud, Pierre Charbonnier, Pierre Lobel, Michel Barlaud,
`Universite de Nice-Sophia Antipolis, (FRANCE)
`
`V-30S
`
`Tomographic Reconstruction of Time-Varying Object from
`Linear Time-Sequential Sampled Projections
`Ying Ha Chiu, Sze Fong Yau, Hong Kong University of Science &
`Technology, (HONGKONG)
`
`V-309
`
`The Reconstruction of Subsurface Property Maps using
`Projection onto Convex Sets
`Alberto Malinvemo, David Rossi, Schlumberger-Doll Research, Michael
`Daniel, MIT, (USA)
`
`V-313
`
`Simultaneous Confidence Intervals for Image
`Reconstruction Problems
`Yong Zhang, Alfred O Hero III, W L Rogers, University of Michigan,
`(USA)
`
`V-317
`
`Iterative Reconstruction of Multidimensional
`Objects Buried in Inhomogeneous Elastic Media
`Tarek Habashy, Schlumberger-Doi/ Research, Eveline Bellegarda,
`IBM T J Watson Research Center, (USA)
`
`V-321
`
`IMAGE CODING I
`
`Optimal Entropy Constrained Scalar Quantization for
`Exponential and Laplacian Random Variables
`Gary J Sullivan, PictureTel Corp, (USA)
`
`Lattice Vector Quantization oflmage Wavelet Coefficient
`Vectors Using a Simplified Form of Entropy Coding
`Andrew Woolf, Glynn Rogers, CS/RO Division of Radiophysics,
`(AUSTRALJA)
`
`V-265
`
`Chairperson: Henrik Sorensen, University of Pennsylvania, (USA)
`
`Discrete Multichannel Orthogonal Transforms
`Ioannis Pitas, Anestis Karasaridis, University o/Thessaloniki,
`(GREECE)
`
`V-269
`
`V-325
`
`Page 5 of 12
`
`

`

`An Edge Classification Based Approach to the Post-Processing V-329
`of Transform Coded Images
`John McDonnell, Robert Shorten, Anthony D Fagan, University College
`Dublin, (IRELAND)
`
`Covariance Matrix Matching for Multi-Spectral
`Image Classification
`Paul J Whitbread, Defence Science and Technology Organisation,
`(AUSTRAUA)
`
`Block Predictive Transform Coding of Still Images
`Jianzhong Huang, Georgia Institute of Technology, Sam Liu, Hewlett(cid:173)
`Packard Laboratories, (USA)
`
`V-333
`
`Tracking Subspace Representations of Face Images
`Hsi-Jung Wu, Dulce Ponceleon, Katherine Wang, James Normile,
`Apple Computer Inc, (USA)
`
`V-385
`
`V-389
`
`V-393
`
`Image Coding with Discrete Cosine Transforms
`using Efficient Energy-Based Adaptive Zonal Filtering
`Alessandro Palau, Gagan Mirchandani, University of Vermont, (USA)
`
`V-337
`
`A New Method for Block Effect Removal
`In Low Bit-Rate Image Compression
`Jiebo Luo, Chang Wen Chen, Kevin J Parker, University of Rochester,
`Thomas S Huang, University of Illinois at Urbana-Champaign, (USA)
`
`V-341
`
`Fast Segmented Image Coding using Weakly Separable Bases
`Wilfried Philips, University of Gent, Charilaos
`Christopoulos, ETROIJRJS-VUB, (BELGIUM)
`
`V-345
`
`Enimatlon of Aliasing Error in Layered Coding System
`Masahiro lwahashi, Nagaoka University of Technology,
`Koichi Ohyama, Graphics Communication Laboratories,
`Noriyoshi Kambayashi, Nagaoka University of Technology, (JAPAN)
`
`V-349
`
`icatlon
`
`Two-Stage Discrete Cosine
`on Image Co
`Neng-Chung H
`Technology, (T.
`
`Design or Subband Filters for I
`Perceptual Cri
`V Ralph Algazi,
`Uni~•ersity of Cali
`
`Subband Coding of Colour Images with Limited Palette Size
`Patrick Waldemar, Tor Audun Ramstad, Norwegian
`Institute of Technology, (NORWAY)
`
`V-353
`
`IMAGE MODELLING AND ANALYSIS
`
`Chairperson: Vinod Chandran, Signal Processing Research Centre,
`Queensland University of Technology, (AUSTRALIA)
`
`Inversion of Large-Support Ill-conditioned Linear Operators
`Using a Markov Model with a Line Process
`Mila Nikolova, Ali Mohammad-Djafari, Jerome ldier, J...SS-ESE,
`(FRANCE)
`
`V-357
`
`V-361
`Estimation of 9-th Order Fractal Dimensions
`Daniele Giusto, University of Cagliari, Stefano Fioravanti, University of
`Genoa, (ITALY)
`
`On the Distribution of the DCT Coefficients
`Thierry Eudc, la3i-LCIA, INSA Rouen, Richard Grisel, JCPJ-l2S2,
`Hocinc Chcrifi, TS/ (CNRS 842), Roland Debrie, la3i-LCIA, INSA Rouen,
`(FRANCEJ
`
`V-365
`
`Self-Similarity Modeling for Interpolation and Extrapolation
`of Multi-Viewpoint Image Sets
`Takeshi Naemura, Hiroshi Harashima, The University of Tokyo, (JAPAN)
`
`V-369
`
`Stochastic Modelling and Estimation ofMultispectral
`Image Data
`Richard Schultz, Robert L Stevenson, University of Notre Dame, (USA)
`
`V-373
`
`Approximate Covariance Functions for Grey-Level Gibbs
`Random Fields
`Ibrahim M Elfadcl, MIT Research Laboratory of Electronics, (USA)
`
`V-377
`
`Cluster-Based Probability Model Applied to Image
`Restoration and Compression
`Kris Popat, Rosalind W Picard, MIT Media Laboratory, (USA)
`
`V-381
`
`Texture Class Assignment In TEXSCALE:
`An Evaluation Study
`Jane You, Unil'ersity of South Australia, Harvey A Cohen,
`La Trobe University, (AUSTRALIA)
`
`A CAD Driven Multlscale Approach to Automated Inspection V-397
`Daniel Tretter, Khalid Khawaja, Charles Bouman, Anthony Maciejewski,
`Purdue University, (USA)
`
`VIDEO CODING I
`
`Chairperson: Kiyoharu Aizawa, University of Tokyo, (JAPAN)
`
`Adaptive Post-Processing Algorithms for Low Bit Rate
`Video Signals
`Tsann-Shyong Liu, Institute of Computer Science, (TAIWAN ROC );
`Nikil Jayant, AT & T Bell Laboratories, (USA)
`
`V-401
`
`V-405
`JD Subband Coder for Very Low Bit Rates
`Weng Leong Chooi, King Ngi Ngan, Monash University, (AUSTRALIA)
`
`Performance Evaluation of Video Coding Schemes
`Working at Very Low Bit Rates
`Laura Conlin, Stefano Battista, CSELT, (ITALY)
`
`Simultaneous 3-D Motion Estimation and Wire-Frame Model
`Adaptation Including Photometric Effects for Knowledge-based
`Video Coding
`G6zdc Bozdagi, Bil/rent University, (TURKEY); A Murat Tekalp,
`University of Rochester, ( USA); Lcvent Onural, Bil/rent University,
`'
`(TURKE~
`
`Rate-Distortion Analysis of Variable Block Size
`VQ-Based Motion Compensated Video Codecs
`Sam Liu, Hewltll-Packard Laboratories, (USA)
`
`Use of Steerable Viewing Window (SVW) to Improve
`the Visual Sensation in Face to Face Teleconferencing
`Liyanage C De Silva, Kiyoharu Aizawa, Mitsutoshi Hatori,
`University of Tokyo, (JAPAN)
`
`V-409
`
`V-413
`
`V-417
`
`V-421
`
`Two-Layered DCT Based Coding Scheme for Recording
`Digital HDTV Signals
`Jae Hyun Kim, Gooman Park, Samsung Electronics Co lid, (KOREA)
`
`V-425
`
`Noise Reduction for MPEG Type of Codec
`Li Yan,AT&TBell Laboratories, (USA)
`
`V-429
`
`The Application of Subband Coding in MPEG for Priorit4ed
`ATM Networks
`Brian DeClccnc, Henrik Sorensen, University of Pennsylvania, (USA)
`
`V-433
`
`An lnvesti.tion of JPEG Image and Video Compression
`Using Parallel Processing
`Gregory W Cook, Edward J Delp, Purdue University, (USA)
`
`V-437
`
`Effective Nearly Lossless Compression of Digital Video
`Sequences via Motion Compensated Filtering
`Balas K Natarjan, Vasudcv Bhaskaran, Hew/ell Packard labs, (USA)
`
`V-441
`
`Page 6 of 12
`
`

`

`IMAGE RESTORATION AND ESTIMATION
`
`IMAGE PROCESSING AND ANALYSIS
`
`Chairperson: Trevor Cole, University of Sydney, (AUSTRAUA)
`
`Blind Deconvolution for Multidimensional Images
`RP Millane, Purdue University, (USA): P J Bones,
`H Jiang, Unfrersity of Canterbury, (NEW ZEALAND)
`
`Iterative Adaptive Ip Restoration of Blurred Images
`Wai Ho Pun, Brian D Jeffs, Brigham Young University, (USA)
`
`Applying Generalised Cross-Validation to Image
`Restoration
`Robert Whatmough, Defence Science and Technology Orga_nisation,
`(AUSTRAUA)
`
`On the Equivalence of Normalized Convolution
`and Normalized Differential Convolution
`Carl-Fredrik Westin, Klas Nordberg, Hans Knutsson,
`Linkiiping University, (SWEDEN)
`
`V-445
`
`V-449
`
`V-453
`
`V-457
`
`Sparse Colour and Grey Scale Image Restoration
`Using a Morphological Method
`Alan Harvey, Royal Melbourne Institute of Technology, Harvey Cohen,
`la Trohe University, (4USTRAUA)
`
`V-461
`
`Image Reconstruction From Zeros of the Z-transform
`Charles Parker, Brenda Satherley, Philip J Bones, University of
`Canterbury, (NEW ZEALAND)
`
`A Subspace Decomposition Method for Point Source
`Localization in Blurred Images
`Melin Gunsay, Brian D Jeffs, Brigham Young University, (USA)
`
`A Fast Parallel Projection Algorithm for Set Theoretic
`Image Recovery
`PL Combettes, H Puh, City University of New York, (USA)
`
`V-465
`
`V-469
`
`V-473
`
`VIDEO CODING II
`
`Chairperson: Murat Tekalp, University of Rochester, (USA)
`
`Frequency Scalable Video Coding Using the MDCT
`Andrew Johnson, Telecom Australia (AUSTRAUA); John Princen,
`AT & T Bell laboratories, (USA); Ming Chan, Telecom Australia,
`(AUSTRAUA)
`
`V-477
`
`V-481
`Hybrid Video Coding for Low Bit-Rate Applications
`Feng-Ming Wang, C-Cube Microsystems, Sam Liu, Hewlett Packard
`laboratories, (USA)
`
`A Novel Tree-Structured Video Coder
`Francesco De Natale, University of Genoa, Giuseppe Desoli,
`PhD Engineering, Daniele Giusto, University of Genoa, (ITALY)
`
`V-485
`
`V-489
`Library-based Image Coding
`Nuno Vasconcelos, Andrew Lippman, MIT Media laboratory, (USA)
`
`Rate and Resolution Scalable Subband Coding of Video
`David Taubman, Avideh Zakhor, University of California at
`Berkeley, (USA)
`
`Subband Video Coding with Temporally
`Adaptive Motion Interpolation
`Jungwoo Lee, Bradley Dickinson, Princeton University, (USA)
`
`Efficient Prediction of Uncovered Background
`in Interframe Coding Using Spatial Extrapolation
`Andre Kaup, Tit Aach, RWTH Aachen, (GERMANY)
`
`Rate Buffered Fractal Video
`David Wilson, Jeremy Nicholls, Donald M Monro, University of
`Bath, (UK)
`
`V-493
`
`V-497
`
`V-501
`
`V-505
`
`Chairperson: W Boles,
`Queensland University of Technology, (AUSTRALIA)
`
`Importance of
`Principal Com
`Tanweer Kabir
`
`Fast Design of Optimal Array Filters
`Phillip Musumeci, Australian Defence Force Academy,
`(AUSTRAUA)
`
`3-D Band-Limitation by Motion Adaptive Spatial Filtering
`Hyun-Soo Kang, Korea Advanced Institute of Science and
`Technology, Jong-Hun Kim, DAEWOO £lee. Co. Ltd, Jung-Hee
`Lee, Seong-Dae Kim, Korea Advanced Institute of Science and
`Technology, (KOREA)
`
`The Automatic Generation of 3D Object Model
`from Range Image
`Wentao Zheng, Hiroshi Harashima, University of Tokyo, (JAPAN)
`
`Real-Time Determination of the Signal-to-Noise
`Ratio of Partly Coherent Seismic Time Series
`Peter K M0ller, Technical University of Denmark and Signal-Data,
`(DENMARK)
`
`A Focusing Algorithm for the Optical Array Imaging System
`Osamu Ikeda, Takushoku University, (JAPAN)
`
`Robust N-Dimenslonal Orientation Estimation
`Using Quadrature Filters and Tensor Whitening
`Hans Knutsson, Mats Anderssom, Linkoping University,
`(SWEDEN)
`
`Global Non-Linear Multigrld Optimization for
`Image Analysis Tasks
`Jean-Marc Laferte, Patrick Perez, Fabrice Heitz, INRIA - Rennes,
`(FRANCE)
`
`V-509
`
`V-513
`
`V-517
`
`V-521
`
`V-525
`
`V-529
`
`V-533
`
`Maximum Likelihood Scale Estimation for a Class of
`Markov Random Fields
`Charles A Bouman, Purdue University, Ken Sauer, University of Notre
`Dame,(USA)
`
`V-537
`
`Divergence Penalty for Image Regularization
`Joseph A O'Sullivan, Washington University in St Louis, (USA)
`
`New Prospects in Line Detection for Remote
`Sensing Images
`Nicolas Meriel, Hebrew University of Jerusalem, (ISRAEL);
`Josiane Zerubia, INRIA, (FRANCE)
`
`Bayesian Restoration of Millimeter Wave Imagery
`BR Hunt, D DeKruger, University of Arizona, (USA)
`
`V-541
`
`V-545
`
`V-549
`
`IMAGE CODING II
`
`Chairperson: Abdesselam Bouzerdoum,
`University of Adelaide, (AUSTRAUA)
`
`Residual Image Coding Using Mathematical Morphology
`Josep R Casas, Luis Torres, Universitat Politecnica de Catalunya,
`(SPAIN)
`
`V-553
`
`Performance Comparison of Different Wavelet Functions in
`Compression and Reconstruction oflmages
`Prakash Kumar Pati, Sumana Gupta, Indian Institute ofTec/mology,
`(IND/A)
`
`•
`
`Fractal Image Compression without Searching
`Donald Monro, Stuart Woolley, University of Bath, (UK)
`
`V-557
`
`Page 7 of 12
`
`

`

`On the Convergence of Fractal Transforms
`Bernd Hiirtgen, Thomas Hain, Aachen University ofTechnology,
`(GERMANY)
`
`New Improved Collage Theorem with Applications to
`Multiresolution Fractal Image Coding
`Geir E 0ien, Rogaland University Centre, (NORWAY); Zachi
`Baharav, Technion - Israel Institute ofTechnology, (ISRAEL);
`Skjalg Lep~y, Consensus Analysis AIS, (NORWAY) ; E Kamin,
`IBM Israel Science and Technology, David Malah, Technion -
`Israel Institute ofTechnology, (ISRAEL)
`
`Overlapped Adaptive Partitioning for Image Coding
`Based on the Theory oflterative Functions Systems
`Emmanuel Reusens, Swiss Federal Institute ofTechnology,
`(SWITZERLAND)
`
`Pusition-Dependent Encoding
`John G Apostolopoulos, Aleksandar Pfajfer, Hae Mook Jung,
`Jae S Lim, Massachusetts Institute of Technology, (USA)
`
`3-D Image Coding Based on Affine Transform
`Toshiaki Fujii, Hiroshi Harashima, University ofTokyo, (JAPAN)
`
`A Study of Convex Coders With An Application to
`Image Coding
`Kohtaro Asai, Mitsubishi Electric Corp, (JAPAN) ; Nguyen T Thao,
`Hong Kong University of Science and Technology, (HONG
`KONG); Manin Vetter Ii, University of Southern California/
`Columbia University, (USA)
`
`V-561
`
`V-565
`
`V-569
`
`V-573
`
`V-S77
`
`V-S81
`
`An Extension to the Analytical Gabor Expansion with
`Applications in Image Coding
`Andreas Teuner, University of Duisburg, (GERMANY); Per Asbeck
`Nielsen, Hedrich J Hosticka, Fraunhofer Institute of
`Microelectronic Circuits and Systems, (GERMANY)
`
`V-585
`
`VECTOR QUANTIZATION
`
`Chairperson: Eve Riskin, University of Washington, (USA)
`
`Image Coding with Overlapped Projection and Pyramid
`Vector Quantization
`Rosa Lancini, CEFRIEL-COGEFO, Emanuele Marconetti,
`Stefano Tubaro, Politecnico di Milano, (!TALY)
`
`Segmentation Based Coding of Textures Using Stochastic
`Vector Quantization
`Luis Torres, Josep R Casas, S de Diego, Universitat Politecnica de
`Catalunya, (SPAIN)
`
`Image Coding using Pyramid Vector Quantization of
`Subband Coefficients
`Ely Tsem, Teresa Meng, Stanford University, (USA)
`
`Vector Quantization over a Noisy Channel Using
`Soft Decision Decoding
`Mikael Skoglund, Per Hedelin, Chalmers University of Technology,
`(SWEDEN)
`
`Predictive Mean Search Algorithm for Vector Quantization
`of Images
`Kwok-Tung Lo, Hong Kong Polytechnic, (HONG KONG);
`Jian Feng, University of New South Wales, (AUSTRALIA)
`
`V-593
`
`V-S97
`
`V-601
`
`V-60S
`
`V-609
`
`An Efficient Neural Prediction for Vector Quantization
`Roberto Fioravanti, PhD Engineering, Genova, Stefano Fioravanti,
`Universita di Genova, Daniele D Giusto, Universita di Cagliari, (!TALY)
`
`V-613
`
`Next-State Functions for Finite-State Vector Quantization
`Nasser M Nasrabadi, Syed A Rizvi, State University of New York
`at Buffalo, (USA)
`
`V-617
`
`Comparison of"Wavelet" Filters and Subband
`Analysis Structures for Still Image Compression
`James Andrew, Philip Ogunbona, Frank Paoloni, University of
`Wollongong, (AUSTRALIA)
`
`V-589
`
`V-621
`Lapped RVQ and Alphabet and Entropy Constraints
`Francesco Nesci, Faouzi Kossentini, Mark J T Smith, Georgia Institute of
`Technology, (USA)
`
`One-Pass Adaptive Universal Vector Quantization
`Michelle Effros, Stanford University, Philip A Chou, Xerox Palo
`Alto Research Center, Robert M Gray, Stanford University, (USA)
`
`Image Reconstruction using Vector Quantized Linear
`Interpolation
`Sheila S Hemami, Robert M Gray, Stanford University, (USA)
`
`V-625
`
`V-629
`
`* Manuscript unavailable for publication.
`Late papers may be included in the rear of Volume 6.
`
`Page 8 of 12
`
`

`

`ON-LINE CURSIVE HANDWRITING RECOGNITION USING SPEECH
`RECOGNITION METHODS
`
`\
`
`Thad Starnert, John l\fakhoul, Richard Schwartz, and George Chou
`
`BBN Systems and Technologies
`70 Fawcett Street, Cambridge, MA 02138
`Email: Makhoul@bbn.com
`
`ABSTRACT
`A hidden Markov m~ del (HMM) based continuous speech
`recognition system is applied to on-line cursive handwrit(cid:173)
`ing recognition. The base system is unmodified except for
`using handwriting feature vectors instead of speech . Due
`to inherent properties of HMMs , segmentation of the hand(cid:173)
`written script sentences is unnecessary. A 1. 1 % word er(cid:173)
`ror rate is achieved for a 3050 word lexicon, 52 character ,
`writer-dependent task and 3%-5% word error rates are ob(cid:173)
`tained for six different writers in a 25,595 word lexicon ,
`86 character, writer-dependent task. Similarities and dif(cid:173)
`ferences between the continuous speech and on-line cursive
`handwriting recognition tasks <l,re explored; the handwrit(cid:173)
`ing database collected over the past year is described· and
`specific implementation details of the handwriting s;stem
`are discussed .
`
`1. INTRODUCTION
`
`Traditionally, the first step in handwriting recognition is
`the segmentation of words into component characters [l).
`However, in modern continuous speech recognition efforts,
`phonemes are not segmented before training or recognition.
`l~stead, segmentation occurs simultaneously with recogni(cid:173)
`tion . If such a system could be adapted for handwriting,
`the very difficult and time consuming issue of segmentation
`could be avoided . This paper presents an approach for the
`~utomatic recognition of on-line cursive handwriting (using
`mput from a pentop computer) by using continuous speech
`recognition methods. Specifically, the u