Lecture Notes in Computer Science 1st Edition by

Springer ISBN pdf download

https://ebookball.com/product/lecture-notes-in-computer-
science-1st-edition-by-springer-isbn-10572/

Explore and download more ebooks or textbooks

at ebookball.com
Here are some recommended products for you. Click the link to
download, or explore more at ebookball.com

Lecture Notes in Computer Science 4075 Lecture Notes in

Bioinformatics 1st edition by Victor Markowitz, Ulf Leser,
Felix Naumann, Barbara Eckman 9783540365952
https://ebookball.com/product/lecture-notes-in-computer-
science-4075-lecture-notes-in-bioinformatics-1st-edition-by-victor-
markowitz-ulf-leser-felix-naumann-barbara-eckman-9783540365952-19912/

Iterative Software Engineering for Multiagent Systems

Lecture Notes in Computer Science 1994 Lecture Notes in
Artificial Intelligence 1st edition by Jürgen Lind ISBN
3540421661 9783540421661
https://ebookball.com/product/iterative-software-engineering-for-
multiagent-systems-lecture-notes-in-computer-science-1994-lecture-
notes-in-artificial-intelligence-1st-edition-by-ja1-4rgen-lind-
isbn-3540421661-9783540421661-19638/

Automata for Branching and Layered Temporal Structures

Lecture Notes in Computer Science 5955 Lecture Notes in
Artificial Intelligence 1st edition by Gabriele Puppis
ISBN 3642118801 978-3642118807
https://ebookball.com/product/automata-for-branching-and-layered-
temporal-structures-lecture-notes-in-computer-science-5955-lecture-
notes-in-artificial-intelligence-1st-edition-by-gabriele-puppis-
isbn-3642118801-978-3642118807-195/

Lecture Notes in Computer Science 5549 1st Edition by

Kenneth Church, Alexander Gelbukh ISBN 3642003818
9783642003813
https://ebookball.com/product/lecture-notes-in-computer-
science-5549-1st-edition-by-kenneth-church-alexander-gelbukh-
isbn-3642003818-9783642003813-19904/
Qualitative Spatial Reasoning with Topological Information
Lecture Notes in Computer Science 2293 Lecture Notes in
Artificial Intelligence 1st edition by Jochen Renz ISBN
3540433465 Â 978-3540433460
https://ebookball.com/product/qualitative-spatial-reasoning-with-
topological-information-lecture-notes-in-computer-
science-2293-lecture-notes-in-artificial-intelligence-1st-edition-by-
jochen-renz-isbn-3540433465-978-3540433460-196/

Lecture Notes in Computer Science 1561 1st edition by

Mihir Bellare, Ivan Bjerre Damgård 9783540489696

https://ebookball.com/product/lecture-notes-in-computer-
science-1561-1st-edition-by-mihir-bellare-ivan-bjerre-damgay-
rd-9783540489696-19954/

The Seventeen Provers of the World Lecture Notes in

Computer Science 3600 Lecture Notes in Artificial
Intelligence 1st edition by Freek Wiedijk, Freek Wiedijk
ISBN 3540307044 Â 978-3540307044
https://ebookball.com/product/the-seventeen-provers-of-the-world-
lecture-notes-in-computer-science-3600-lecture-notes-in-artificial-
intelligence-1st-edition-by-freek-wiedijk-freek-wiedijk-
isbn-3540307044-978-3540307044-19614/

Lecture Notes in Computer Science 1st edition by Gerard

Cornuejols , Rainer Burkard , Gerhard Woeginger ISBN
3540660194Â 9783540660194
https://ebookball.com/product/lecture-notes-in-computer-science-1st-
edition-by-gerard-cornuejols-rainer-burkard-gerhard-woeginger-
isbn-3540660194-9783540660194-19670/

Lecture Notes in Computer Science 1st edition by Gerard

Cornuejols , Rainer Burkard , Gerhard WoegingerISBN
3540660194Â 9783540660194
https://ebookball.com/product/lecture-notes-in-computer-science-1st-
edition-by-gerard-cornuejols-rainer-burkard-gerhard-
woegingerisbn-3540660194-9783540660194-19668/
Lecture Notes in Computer Science 5416
Commenced Publication in 1973
Founding and Former Series Editors:
Gerhard Goos, Juris Hartmanis, and Jan van Leeuwen

Editorial Board
David Hutchison
Lancaster University, UK
Takeo Kanade
Carnegie Mellon University, Pittsburgh, PA, USA
Josef Kittler
University of Surrey, Guildford, UK
Jon M. Kleinberg
Cornell University, Ithaca, NY, USA
Alfred Kobsa
University of California, Irvine, CA, USA
Friedemann Mattern
ETH Zurich, Switzerland
John C. Mitchell
Stanford University, CA, USA
Moni Naor
Weizmann Institute of Science, Rehovot, Israel
Oscar Nierstrasz
University of Bern, Switzerland
C. Pandu Rangan
Indian Institute of Technology, Madras, India
Bernhard Steffen
University of Dortmund, Germany
Madhu Sudan
Massachusetts Institute of Technology, MA, USA
Demetri Terzopoulos
University of California, Los Angeles, CA, USA
Doug Tygar
University of California, Berkeley, CA, USA
Gerhard Weikum
Max-Planck Institute of Computer Science, Saarbruecken, Germany
Frank Nielsen (Ed.)

Emerging Trends
in Visual Computing

LIX Fall Colloquium, ETVC 2008

Palaiseau, France, November 18-20, 2008
Revised Invited Papers

13
Volume Editor

Frank Nielsen
Ecole Polytechnique, LIX
Route de Saclay, 91128 Palaiseau Cedex, France
E-mail: [email protected]

and

Sony Computer Science Laboratories, Inc.

3-14-13 Higashi Gotanda 3F, 141-0022 Shinagawa-ku, Tokyo, Japan
E-mail: [email protected]

Library of Congress Control Number: Applied for

CR Subject Classification (1998): I.4, I.5, I.2.10, I.3.3, I.3.5, I.3.7, I.2.6, F.2, G.1.2

LNCS Sublibrary: SL 6 – Image Processing, Computer Vision, Pattern Recognition,

and Graphics

ISSN 0302-9743
ISBN-10 3-642-00825-9 Springer Berlin Heidelberg New York
ISBN-13 978-3-642-00825-2 Springer Berlin Heidelberg New York

This work is subject to copyright. All rights are reserved, whether the whole or part of the material is
concerned, specifically the rights of translation, reprinting, re-use of illustrations, recitation, broadcasting,
reproduction on microfilms or in any other way, and storage in data banks. Duplication of this publication
or parts thereof is permitted only under the provisions of the German Copyright Law of September 9, 1965,
in its current version, and permission for use must always be obtained from Springer. Violations are liable
to prosecution under the German Copyright Law.
springer.com
© Springer-Verlag Berlin Heidelberg 2009
Printed in Germany
Typesetting: Camera-ready by author, data conversion by Scientific Publishing Services, Chennai, India
Printed on acid-free paper SPIN: 12612574 06/3180 543210
 Preface

ETVC 2008, the fall colloquium of the computer science department (LIX) of the
École Polytechnique, held in Palaiseau, France, November 18-20, 2008, focused
on the Emerging Trends in Visual Computing. The colloquium gave scientists the
opportunity to sketch a state-of-the-art picture of the mathematical foundations
of visual computing.
We were delighted to invite and welcome the following distinguished speakers
to ETVC 2008 (listed in alphabetical order):

– Shun-ichi AMARI (Mathematical Neuroscience Laboratory, Brain Science

Institute, RIKEN, Wako-Shi, Japan): Information Geometry and Its
Applications
– Tetsuo ASANO (School of Information Science, Japan Advanced Institute
of Science and Technology, JAIST, Japan): Constant-Working-Space Algo-
rithms for Image Processing
– Francis BACH (INRIA/ENS, France): Machine Learning and Kernel Meth-
ods for Computer Vision
– Frédéric BARBARESCO (Thales Air Systems, France): Applications of In-
formation Geometry to Radar Signal Processing
– Michel BARLAUD (I3S CNRS, University of Nice-Sophia-Antipolis, Poly-
tech’Nice & Institut Universitaire de France, France): Image Retrieval via
Kullback Divergence of Patches of Wavelets Coefficients in the k-NN
Framework
– Jean-Daniel BOISSONNAT (GEOMETRICA, INRIA Sophia-Antipolis,
France): Certified Mesh Generation
– Pascal FUA (EPFL, CVLAB, Switzerland): Recovering Shape and Motion
from Video Sequences
– Markus GROSS (Department of Computer Science, Institute of Scientific
Computing, Swiss Federal Institute of Technology Zurich, ETHZ, Switzer-
land): 3D Video: A Fusion of Graphics and Vision
– Xianfeng David GU (State University of New York at Stony Brook, USA):
Discrete Curvature Flow for Surfaces and 3-Manifolds
– Leonidas GUIBAS (Computer Science Department, Stanford University,
USA): Detection of Symmetries and Repeated Patterns in 3D Point Cloud
Data
– Sylvain LAZARD (VEGAS, INRIA LORIA Nancy, France): 3D Visibility
and Lines in Space
VI Preface

– Stéphane MALLAT (École Polytechnique, Centre de Mathématiques Ap-

pliquées (CMAP), France): Sparse Geometric Super-Resolution
– Hiroshi MATSUZOE (Department of Computer Science and Engineering,
Graduate School of Engineering, Nagoya Institute of Technology, NITECH,
Japan): Computational Geometry from the Viewpoint of Affine Differential
Geometry
– Dimitris METAXAS (Computational Biomedicine Imaging and Modeling
Center, CBMI, Rutgers University, USA): Unifying Subspace and Distance
Metric Learning with Bhattacharyya Coefficient for Image Classification
– Frank NIELSEN (LIX, École Polytechnique, Paris, France & Sony Com-
puter Science Laboratories Inc., Tokyo, Japan): Computational Geometry in
Dually Flat Spaces: Theory, Applications and Perspectives
– Richard NOCK (CEREGMIA, University of Antilles-Guyane, France): The
Intrinsic Geometries of Learning
– Nikos PARAGIOS (École Centrale de Paris, ECP, Paris, France): Procedural
Modeling of Architectures: Towards Large Scale Visual Reconstruction
– Xavier PENNEC (ASCLEPIOS, INRIA Sophia-Antipolis, France): Statis-
tical Computing on Manifolds for Computational Anatomy
– Ramesh RASKAR (MIT Media Lab, USA): Computational Photography:
Epsilon to Coded Imaging
– Cordelia SCHMID (LEAR, INRIA Grenoble, France): Large-Scale Object
Recognition Systems
– Gabriel TAUBIN (Division of Engineering, Brown University, USA): Shape
from Depth Discontinuities
– Baba VEMURI (CISE Dept., University of Florida, USA): Information-
Theoretic Algorithms for Diffusion Tensor Imaging
– Suresh VENKATASUBRAMANIAN (School of Computing, University of
Utah, USA): Non-standard Geometries and Data Analysis
– Martin VETTERLI (School of Computer and Communication Sciences,
EPFL, Switzerland): Sparse Sampling: Variations on a Theme by Shannon
– Jun ZHANG (Department of Psychology, University of Michigan, USA):
Information Geometry: Duality, Convexity and Divergences
Invited speakers were encouraged to submit a state-of-the-art chapter on their
research area. The review process was carried out by members of the Program
Committee and other reviewers. We would like to sincerely thank the contribut-
ing authors and thank the reviewers for the careful feedback that helped the
authors prepare their camera-ready papers.
Videos of the lectures synchronized with slides are available from
www.videolectures.net
 Preface VII

We were very pleased to welcome all the 150+ participants to ETVC 2008.
For those who did not attend, we hope the chapters of this publication provide
a good snapshot of the current research status in visual computing.

December 2008 Frank Nielsen

Group picture of the participants at ETVC 2008 (November 19, 2008)

 Organization

Frank Nielsen (Program Chair)

Evelyne Rayssac (Secretary)

Corinne Poulain (Secretary)

Philippe Baptiste (Financial Advisor)

Jean-Marc Steyaert (Scientific Advisor)

Luca Castelli Aleardi (Photographer)

Referees
S. Boltz R. Keriven S. Owada
F. Chazal F. Nielsen M. Pauly
B. Lévy R. Nock A. Vigneron
A. André T. Nakamura
F. Hetroy S. Oudot

Sponsoring Institutions
We gratefully acknowledge the following institutions for their generous support:

– CNRS
– DIGITEO
– École Polytechnique
– Groupe de Recherche Informatique & Mathématique (GdR IM)
– University of Antilles-Guyane, CEREGMIA Department
 Table of Contents

Geometric Computing
Abstracts of the LIX Fall Colloquium 2008: Emerging Trends in Visual
Computing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
Frank Nielsen

From Segmented Images to Good Quality Meshes Using Delaunay

Refin ement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
Jean-Daniel Boissonnat, Jean-Philippe Pons, and Mariette Yvinec

Information Geometry and Applications

Discrete Curvature Flows for Surfaces and 3-Manifolds . . . . . . . . . . . . . . . 38
Xiaotian Yin, Miao Jin, Feng Luo, and Xianfeng David Gu

Information Geometry and Its Applications: Convex Function and

Dually Flat Manifold . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
Shun-ichi Amari

Computational Geometry from the Viewpoint of Affine Differential

Geometry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103
Hiroshi Matsuzoe

Interactions between Symmetric Cone and Information Geometries:

Bruhat-Tits and Siegel Spaces Models for High Resolution
Autoregressive Doppler Imagery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 124
Frederic Barbaresco

Clustering Multivariate Normal Distributions . . . . . . . . . . . . . . . . . . . . . . . . 164

Frank Nielsen and Richard Nock

Computer Graphics and Vision

Intrinsic Geometries in Learning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 175
Richard Nock and Frank Nielsen

Shape from Depth Discontinuities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 216

Gabriel Taubin, Daniel Crispell, Douglas Lanman, Peter Sibley, and
Yong Zhao

Computational Photography: Epsilon to Coded Photography . . . . . . . . . . 238

Ramesh Raskar
XII Table of Contents

Unifying Subspace and Distance Metric Learning with Bhattacharyya

Coefficient for Image Classification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 254
Qingshan Liu and Dimitris N. Metaxas

Information Retrieval
Constant-Working-Space Algorithms for Image Processing . . . . . . . . . . . . . 268
Tetsuo Asano

Sparse Multiscale Patches for Image Processing . . . . . . . . . . . . . . . . . . . . . . 284

Paolo Piro, Sandrine Anthoine, Eric Debreuve, and Michel Barlaud

Medical Imaging and Computational Anatomy

Recent Advances in Large Scale Image Search . . . . . . . . . . . . . . . . . . . . . . . 305
Herve Jegou, Matthijs Douze, and Cordelia Schmid

Information Theoretic Methods for Diffusion-Weighted MRI Analysis . . . 327

Angelos Barmpoutis and Baba C. Vemuri

Statistical Computing on Manifolds: From Riemannian Geometry to

Computational Anatomy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 347
Xavier Pennec

Author Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 387

 Abstracts of the LIX Fall Colloquium 2008:
Emerging Trends in Visual Computing

Frank Nielsen

Ecole Polytechnique, Palaiseau, France

Sony CSL, Tokyo, Japan

Abstract. We list the abstracts of the distinguished speakers that par-

ticipated to the 2008 LIX fall colloquium.

Leonidas GUIBAS
Computer Science Department, Stanford University, USA

Detection of Symmetries and Repeated Patterns in 3D Point Cloud Data

Digital models of physical shapes are becoming ubiquitous in our economy and
life. Such models are sometimes designed ab initio using CAD tools, but more
and more often they are based on existing real objects whose shape is acquired
using various 3D scanning technologies. In most instances, the original scanner
data is just a set, but a very large set, of points sampled from the surface of
the object. We are interested in tools for understanding the local and global
structure of such large-scale scanned geometry for a variety of tasks, including
model completion, reverse engineering, shape comparison and retrieval, shape
editing, inclusion in virtual worlds and simulations, etc. This talk will present a
number of point-based techniques for discovering global structure in 3D data sets,
including partial and approximate symmetries, shared parts, repeated patterns,
etc. It is also of interest to perform such structure discovery across multiple data
sets distributed in a network, without actually ever bring them all to the same
host.

Xianfeng David GU
State University of New York at Stony Brook, USA

Discrete Curvature Flow for Surfaces and 3-Manifolds

This talk introduce the concepts, theories and algorithms for discrete curvature
flows for surfaces with arbitrary topologies. Discrete curvature flow for hyperbolic
3-manifolds with geodesic boundaries are also explained. Curvature flow method
can be used to design Riemannian metrics by prescribed curvatures, and applied
for parameterization in graphics, shape registration and comparison in vision
and brain mapping in medical imaging, spline construction in computer aided
geometric design, and many other engineering fields.

F. Nielsen (Ed.): ETVC 2008, LNCS 5416, pp. 1–12, 2009.


c Springer-Verlag Berlin Heidelberg 2009
2 F. Nielsen

Jean-Daniel BOISSONNAT
GEOMETRICA, INRIA Sophia-Antipolis, France

Certified Mesh Generation

Given a domain D, the problem of mesh generation is to construct a simplicial
complex that approximates D in both a topological and a geometrical sense and
whose elements satisfy various constraints such as size, aspect ratio or anisotropy.
The talk will cover some recent results on triangulating surfaces and volumes by
Delaunay refinement, anisotropic mesh generation and surface reconstruction.
Applications in medical images, computer vision and geology will be discussed.

Baba VEMURI
CISE Dept., University of Florida, USA

Information-Theoretic Algorithms for Diffusion Tensor Imaging

Concepts from Information Theory have been used quite widely in Image
Processing, Computer Vision and Medical Image Analysis for several decades
now. Most widely used concepts are that of KL-divergence, minimum descrip-
tion length (MDL), etc. These concepts have been popularly employed for image
registration, segmentation, classification etc. In this chapter we review several
methods, mostly developed by our group at the Center for Vision, Graphics &
Medical Imaging in the University of Florida, that glean concepts from Informa-
tion Theory and apply them to achieve analysis of Diffusion-Weighted Magnetic
Resonance (DW-MRI) data. This relatively new MRI modality allows one to
non-invasively infer axonal connectivity patterns in the central nervous system.
The focus of this chapter is to review automated image analysis techniques that
allow us to automatically segment the region of interest in the DWMRI im-
age wherein one might want to track the axonal pathways and also methods
to reconstruct complex local tissue geometries containing axonal fiber crossings.
Implementation results illustrating the algorithm application to real DW-MRI
data sets are depicted to demonstrate the effectiveness of the methods reviewed.

Xavier PENNEC
ASCLEPIOS, INRIA Sophia-Antipolis, France

Statistical Computing on Manifolds for Computational Anatomy

Computational anatomy is an emerging discipline that aims at analyzing and
modeling the individual anatomy of organs and their biological variability across a
population. The goal is not only to model the normal variations among a popula-
tion, but also discover morphological differences between normal and pathological
populations, and possibly to detect, model and classify the pathologies from struc-
tural abnormalities. Applications are very important both in neuroscience, to min-
imize the influence of the anatomical variability in functional group analysis, and in
medical imaging, to better drive the adaptation of generic models of the anatomy
(atlas) into patient-specific data (personalization). However, understanding and
 Abstracts of the LIX Fall Colloquium 2008 3

modeling the shape of organs is made difficult by the absence of physical models
for comparing different subjects, the complexity of shapes, and the high number
of degrees of freedom implied. Moreover, the geometric nature of the anatomical
features usually extracted raises the need for statistics and computational meth-
ods on objects that do not belong to standard Euclidean spaces. We investigate in
this chapter the Riemannian metric as a basis for developing generic algorithms
to compute on manifolds. We show that few computational tools derived from this
structure can be used in practice as the atoms to build more complex generic algo-
rithms such as mean computation, Mahalanobis distance, interpolation, filtering
and anisotropic diffusion on fields of geometric features. This computational frame-
work is illustrated with the joint estimation and anisotropic smoothing of diffusion
tensor images and with the modeling of the brain variability from sulcal lines.

Cordelia SCHMID
LEAR, INRIA Grenoble, France

Large-Scale Object Recognition Systems

This paper introduces recent methods for large scale image search. State-of-the-
art methods build on the bag-of-features image representation. We first analyze
bag-of-features in the framework of approximate nearest neighbor search. This
shows the sub-optimality of such a representation for matching descriptors and
leads us to derive a more precise representation based on 1) Hamming embedding
(HE) and 2) weak geometric consistency constraints (WGC). HE provides binary
signatures that refine the matching based on visual words. WGC filters matching
descriptors that are not consistent in terms of angle and scale. HE and WGC are
integrated within the inverted file and are efficiently exploited for all images, even
in the case of very large datasets. Experiments performed on a dataset of one
million of images show a significant improvement due to the binary signature and
the weak geometric consistency constraints, as well as their efficiency. Estimation
of the full geometric transformation, i.e., a re-ranking step on a short list of
images, is complementary to our weak geometric consistency constraints and
allows to further improve the accuracy.

Pascal FUA
EPFL, CVLAB, Swiss

Recovering Shape and Motion from Video Sequences

In recent years, because cameras have become inexpensive and ever more preva-
lent, there has been increasing interest in video-based modeling of shape and
motion. This has many potential applications in areas such as electronic pub-
lishing, entertainment, sports medicine and athletic training. It, however, is an
inherently difficult task because the image-data is often incomplete, noisy, and
ambiguous. In our work, we focus on the recovery of deformable and articulated
3D motion from single video sequences. In this talk, I will present the models we
4 F. Nielsen

have developed for this purpose and demonstrate the applicability of our tech-
nology for Augmented Reality and human body tracking purposes. Finally, I will
present some open research issues and discuss our plans for future developments.

Ramesh RASKAR
MIT Media Lab, USA

Computational Photography: Epsilon to Coded Imaging

Computational photography combines plentiful computing, digital sensors, mod-
ern optics, actuators, and smart lights to escape the limitations of traditional
cameras, enables novel imaging applications and simplifies many computer vision
tasks. However, a majority of current Computational Photography methods in-
volve taking multiple sequential photos by changing scene parameters and fusing
the photos to create a richer representation. The goal of Coded Computational
Photography is to modify the optics, illumination or sensors at the time of cap-
ture so that the scene properties are encoded in a single (or a few) photographs.
We describe several applications of coding exposure, aperture, illumination and
sensing and describe emerging techniques to recover scene parameters from coded
photographs.

Dimitris METAXAS
Computational Biomedicine Imaging and Modeling Center, CBMI, Rutgers Uni-
versity, USA

Unifying Subspace and Distance Metric Learning with Bhattacharyya Coefficient

for Image Classification
In this talk, we propose a unified scheme of subspace and distance metric learning
under the Bayesian framework for image classification. According to the local
distribution of data, we divide the k-nearest neighbors of each sample into the
intra-class set and the inter-class set, and we aim to learn a distance metric in
the embedding subspace, which can make the distances between the sample and
its intra-class set smaller than the distances between it and its inter-class set. To
reach this goal, we consider the intra-class distances and the inter-class distances
to be from two different probability distributions respectively, and we model the
goal with minimizing the overlap between two distributions. Inspired by the
Bayesian classification error estimation, we formulate the objective function by
minimizing the Bhattachyrra coefficient between two distributions. We further
extend it with the kernel trick to learn nonlinear distance metric. The power and
generality of the proposed approach are demonstrated by a series of experiments
on the CMU-PIE face database, the extended YALE face database, and the
COREL-5000 nature image database.

Nikos PARAGIOS
Ecole Centrale de Paris, ECP, Paris, France
 Abstracts of the LIX Fall Colloquium 2008 5

Procedural Modeling of Architectures: Towards Large Scale Visual Reconstruction

Three-dimensional content is a novel modality used in numerous domains like
navigation, post production & cinematography, architectural modeling and ur-
ban planning. These domains have benefited from the enormous progress has
been made on 3D reconstruction from images. Such a problem consists of build-
ing geometric models of the observed environment. State of the art methods can
deliver excellent results in a small scale but suffer from being local and cannot be
considered in a large scale reconstruction process since the assumption of recov-
ering images from multiple views for an important number of buildings is rather
unrealistic. On the other hand several efforts have been made in the graphics
community towards content creation with city engines. Such models are purely
graphics-based and given a set of rules (grammars) as well as dictionary of ar-
chitectures (buildings) can produce virtual cities. Such engines could become far
more realistic through the use of actual city models as well as knowledge of build-
ing architectures. Developing 3D models/rules/grammars that are image-based
and coupling these models with actual observations is the greatest challenge of
urban modeling. Solving the large-scale geometric modeling problem from min-
imal content could create novel means of world representation as well as novel
markets and applications. In this talk, we will present some preliminary results
on large scale modeling and reconstruction through architectural grammars.

Gabriel TAUBIN
Division of Engineering, Brown University, USA

Shape from Depth Discontinuities

We propose a new primal-dual framework for representation, capture, processing,
and display of piecewise smooth surfaces, where the dual space is the space of
oriented 3D lines, or rays, as opposed to the traditional dual space of planes.
An image capture process detects points on a depth discontinuity sweep from a
camera moving with respect to an object, or from a static camera and a moving
object. A depth discontinuity sweep is a surface in dual space composed of the
time-dependent family of depth discontinuity curves span as the camera pose
describes a curved path in 3D space. Only part of this surface, which includes
silhouettes, is visible and measurable from the camera. Locally convex points
deep inside concavities can be estimated from the visible non-silhouette depth
discontinuity points. Locally concave point laying at the bottom of concavities,
which do not correspond to visible depth discontinuities, cannot be estimated,
resulting in holes in the reconstructed surface. A first variational approach to
fill the holes, based on fitting an implicit function to a reconstructed oriented
point cloud, produces watertight models.We describe a first complete end-to-end
system for acquiring models of shape and appearance.We use a single multi-flash
camera and turntable for the data acquisition and represent the scanned objects
as point clouds, with each point being described by a 3-D location, a surface
normal, and a Phong appearance model.
6 F. Nielsen

Shun-ichi AMARI
Mathematical Neuroscience Laboratory, Brain Science Institute, RIKEN, Wako-
Shi, Japan

Information Geometry and Its Applications

Information geometry emerged from studies on invariant properties of a manifold
of probability distributions. It includes convex analysis and its duality as a spe-
cial but important part. Here, we begin with a convex function, and construct a
dually flat manifold. The manifold possesses a Riemannian metric, two types of
geodesics, and a divergence function. The generalized Pythagorean theorem and
dual projections theorem are derived therefrom.We construct alpha-geometry,
extending this convex analysis. In this review, geometry of a manifold of proba-
bility distributions is then given, and a plenty of applications are touched upon.
Appendix presents an easily understable introduction to differential geometry
and its duality.

Jun ZHANG
Department of Psychology, University of Michigan, USA

Information Geometry: Duality, Convexity and Divergences

In this talk, I explore the mathematical relationships between duality in in-
formation geometry, convex analysis, and divergence functions. First, from the
fundamental inequality of a convex function, a family of divergence measures
can be constructed, which specializes to the familiar Bregman divergence, Jen-
son difference, beta-divergence, and alpha-divergence, etc. Second, the mixture
parameter turns out to correspond to the alpha ¡-¿ -alpha duality in informa-
tion geometry (which I call “referential duality”, since it is related to the choice
of a reference point for computing divergence). Third, convex conjugate oper-
ation induces another kind of duality in information geometry, namely, that
of biorthogonal coordinates and their transformation (which I call “representa-
tional duality”, since it is related to the expression of geometric quantities, such
as metric, affine connection, curvature, etc of the underlying manifold). Under
this analysis, what is traditionally called “+1/-1 duality” and “e/m duality”
in information geometry reflect two very different meanings of duality that are
nevertheless intimately interwined for dually flat spaces.

Hiroshi MATSUZOE
Department of Computer Science and Engineering Graduate School of Engineer-
ing, Nagoya Institute of Technology, NITECH, Japan

Computational Geometry from the Viewpoint of Affine Differential Geometry

Incidence relations (configurations of vertexes, edges, etc.) are important in
computational geometry. Incidence relations are invariant under the group of
affine transformations. On the other hand, affine differential geometry is to
study hypersurfaces in an affine space that are invariant under the group of
Random documents with unrelated
content Scribd suggests to you:
Flower, ¾ inch in diameter, of 5 to 9 showy petal-like sepals, usually
deep red, occasionally yellow, forms a shallow cup around the
numerous stamens and a conspicuous group of pistils which, after
the flower fades, become a round thimble-shaped seed cluster. The
pedicels, bearing the solitary flowers at their tips, are several inches
long and covered with silky hairs. Plant is about 1 foot tall, with
subdivided leaves near the base and on the sparingly branched
stems. Grows in partial shade in montane zone. Blooms June-July.

This globe anemone, related to the better known pasque flower, is

one of the many less conspicuous plants that add to the charm of a
flowery hillside, yet reserve their more delicate beauty for those who
take time to prowl. This particular specimen was found in a glade
filled with columbines. We would probably not have seen it if we had
not stopped to try one more columbine picture! Anemone canadensis
is a somewhat larger plant with pure white flowers, rather woody
stems and deep green foliage. It grows in shady places along foothill
streams, but only where conditions are to its liking. In these spots it
forms rather dense colonies.

21
 Buttercup Family
Nelson’s Larkspur, Delphinium nelsonii, GREENE
Flowers, ½ inch or more wide, are formed of 5 showy, dark blue,
irregularly shaped sepals, enclosing at their base 4 much smaller
petals of lighter color. The uppermost sepal extends backward as a
slender spur ½ inch or more in length. About a dozen flowers on
slender pedicels group around a central erect stem to form a loose
raceme which often nods slightly at the top. Plant is 10 to 15 inches
tall and bears rather few leaves each sub-divided into linear
segments. Grows in foothills zone. Blooms late April to early June.

This small larkspur of the early spring looks much like the single
larkspur of an old-fashioned garden. Its favorite location is near the
base of a clump of scrub oak where a little snow has drifted in the
winter giving that spot a bit of extra water. The intense blue of these
flowers contrasts well with the leather brown color of last season’s
oak leaves. When spring is farther advanced other taller larkspurs,
such as Delphinium geyeri, called poison-weed by the stockmen,
make a more spectacular showing on low foothills and plains. All of
the larkspurs contain an alkaloid poison which is deadly to cattle and
somewhat dangerous to other stock.

22
 Buttercup Family
Snow Buttercup, Ranunculus adoneus, GRAY

Flowers are an inch across, formed of several (3 to 15) broad,

overlapping golden petals having the glossy sheen of butter. The
sparse leaves are divided into linear lobes. These and the succulent
stems grow a few inches tall, breaking out of frosty soil with flower
bud ready to open. Grows on alpine and sub-alpine slopes near snow
banks. Blooms when snow melts, usually June to early July.
The hardiness of the snow buttercup is its outstanding characteristic.
It comes up through the snow because in the high altitude in which it
lives its time for fruition is short. It pushes a stout knuckle of stem
through the snow crust, attracting the sun’s heat by the dark color of
its stem, then the knuckle straightens, lifting the already formed bud
into an erect position. The bud opens rapidly and proceeds to spread
out in the hole caused by melting. Of the many glossy members of
the buttercup family, there are few of so rich a yellow, or which give
such an appearance of being all flower with inconsiderable leaf and
stem.

23
 Buttercup Family
Globeflower, Trollius laxus, SALISB.

Flower is 1¼ inches across of 5 to 10 (or more) pale cream petal-like

sepals, with numerous yellow stamens and several pistils in the
center. Numerous petals, so dwarfed as hardly to be noticed,
surround the base of the stamens. Plants, 8 to 15 inches tall, often
grow in groups and bear several flowers, each on its own slender
stem. Leaves are dark green and deeply cut into 5 or more spreading
lobes (palmate). Grows in moist rich soil in sub-alpine and alpine
zones. Blooms late May-July.

When the snowbanks melt in the alpine country, hundreds of

temporary runlets carry the snow water to timberline lakes and to
permanent streams. In the wet soil along these runlets and near
these lakes, globeflower is one of the common and very good looking
plants. Both its foliage and its flowers are graceful and charming.
Associated with it is usually marsh marigold, Caltha rotundifolia,
which is also a member of the buttercup family. Our Colorado marsh
marigold is not gold at all, but white—even a bluish-white. It grows
with its feet right in the water. Its leaves are entire and are all at the
base of the sturdy low plant. Its flowers are as large or slightly larger
than those of globeflower. It makes an effective companion for its
more dainty relative.

24
 Buttercup Family
Columbine, Aquilegia coerulea, JAMES
The flower is formed of 5 sepals and 5 petals, alternately arranged
and all of them showy. The sepals are deep blue or sometimes quite
pale, forming a wide saucer-like star 3 inches across; the petals form
a white inner cup 1¾ inches across, and stretch back between the
sepals as hollow, slender 2-inch spurs. Plants are 2 feet or more high
of several delicate stems, usually carrying at their tops numerous
flowers. The deeply cut leaves are mainly concentrated at the plant
base. Grows in rich soil in montane zone, but extends into foothills
and up to timberline. Blooms June-July.

Colorado’s queenly state flower speaks for itself much more

eloquently than humans can speak for it. No portrait can do it justice.
We have found it in the very glade near Palmer Lake where James
first saw it and named it coerulea for its celestial blue. We have found
it in countless aspen groves of the montane zone and finally on rocky
scree near timberline (a more compact plant there—with flowers
sometimes white or of a rosy hue). Always there is the thrill of real
discovery—a new realization of its beauty. A less common and even
more exciting find is the dwarf columbine, Aquilegia saximontana,
that grows between rocks above timberline.

25
 Poppy Family
Prickly Poppy, Argemone intermedia, SWEET
Flower, 3 inches or more across, is formed of 6 brilliant white, paper-
like petals, surrounding numerous golden stamens with, at the very
center, a dark or even black stigma. Blossoms, in loose clusters
opening over a long period, crowd each other slightly at the tops of
the branching stems. Plant is 2 to 5 feet tall, with gray-green leaves
divided into lobes, and with yellowish spines along the stems and leaf
ribs. Grows in plains, foothills and lower montane zones. Blooms May-
September.

These big coarse plants, which may be seen in small groups along
our roads at culvert ends and in neglected fence rows, could be taken
for some sort of thistle if it were not for the amazing flowers which
they display in successive crops throughout the whole summer. The
blossoms look like big circles of white crepe paper with a center of
spun gold. As the season advances, the plants get ragged, but even
in September a few fresh flowers will appear. Some resemblance can
be seen between these blossoms and the Oriental poppies of our
gardens, but only by study of their botanical structure can we find
why they are put in the same family with golden smoke, Corydalis
aurea, of our foothills, and the bleeding-heart of old-fashioned
gardens.

26
 Mustard Family
Wallflower, Erysimum asperum, DC.
Flowers, ½ inch in diameter, are formed of 4 petals arranged like a
Maltese cross, yellow to orange in color. They are clustered into a
round terminal head, the lower flowers of which open first so that
usually tubular seed pods (siliques) have formed near the base by the
time the top of the cluster is in bloom. Plants are 8 inches or more
high, of several stems from one root crown. Grows in foothills,
extending down to plains and up through montane zone. Blooms
May-July.

The mustards are legion. Fields of them add a yellow note to many
western hillsides. They range from weedy poor relations, like
shepherd’s purse, to tall, showy spikes of prince’s plume, Stanleya
apinnata. Wallflower—despite its name suggesting a colorless
personality—is one of the handsome children of the family. Its
flowers, larger than most mustards, range in color from pale yellow,
through orange, to rich bronze shades. By no means all of the
mustards are yellow. The flowers of many of them are white, some,
like the cardamine that grows in abundance along sub-alpine water
runs, being a very showy, brilliant white.

27
 Saxifrage Family
Snowball Saxifrage, Saxifraga rhomboidea, GREENE
Individual flowers are ¼ inch or less across, each with 5 white petals,
and are grouped in a compact, round-topped head about 1 inch in
diameter which forms the top of a naked stem (scape). This scape
rises to a height of 8 inches, or sometimes much less, from the
center of a flat circle of oblong, leathery leaves. As the blossoms age,
the flower cluster becomes loose and sprangly. Grows on moist
slopes in sub-alpine and montane zones. Blooms May-July.

Saxifrage is another large family of quite varied sorts. Gooseberries

and mock orange come within its membership. The numerous species
of alum root, Heuchera, are also included, as are many little alpine
and sub-alpine plants that grow out of rock crevices in our high
mountains. Purple saxifrage, Saxifraga jamesii, with quite large red-
purple flowers, and dotted saxifrage, Saxifraga austromontana, with
tiny white flowers covered with pale dots, are among the best. All of
these seem able to thrive on only a teaspoonful of soil in a rock
crack, if only there is local moisture. The structural features that
bring all these plants within one family are not obvious. The leaves of
many of them are similar to the leaves of a gooseberry bush, though
in some this resemblance is remote, and in others entirely absent.

28
 Orpine Family
Queen’s Crown, Sedum rhodanthum, GRAY

Individual flowers, ¼ inch across, are formed of 4 or 5 bright rose

petals; numerous flowers being congested in a round head an inch or
more in diameter terminating a leafy shoot, several of which rise from
a woody root crown. Plant is 6 to 10 inches high, with narrow, gray-
green, fleshy leaves crowded along the succulent stems. Grows in
wet places alpine and sub-alpine zones. Blooms June-August.
Along the cold, mountain stream trickling out from Lake Isabelle, or
near any similar alpine lake or tarn, grows the Sedum, named
queen’s crown for the rosy-pink crowns of blossoms. These plants like
to have their feet in the water and often help to make the hillocky
mounds on the lake’s edge. Nearby and tolerating drier ground, is the
king’s crown, Sedum integrifolium, with its flatter head of deep
maroon flowers resembling the old-fashioned Bohemian garnet
jewelry. The stems and leaves of these sedums color brilliantly with
the first frosts and add richness to the Persian carpets of timberline in
late August and early September.

29
 Rose Family
Bush Cinquefoil, Potentilla fruticosa, L.

Flowers are an inch in diameter, of 5 broad, golden petals

surrounding 20 or more stamens. Groups of several flowers are borne
at the ends of the numerous short branches. Plant is a dense shrub
about 3 to 4 feet high with many dark, woody, freely-branching
stems. Leaves are pinnate, with usually 5 or 7 narrow linear leaflets.
Grows in moist parts of the montane zone, also in the upper foothills
and the lower sub-alpine zones. Blooms continuously May to
September.

This thornless yellow rose is one of the most widespread and most
ornamental shrubs of mountain areas. Individual clumps are rarely
fully covered with bloom at any one time, tending rather to bring out
a few fresh flowers each day of the season so that all summer long
there are buds, fresh blossoms, groups of faded petals, and small,
dry, fuzzy seeds (achenes) distributed over the plant. Other species
of Potentilla grow also in our mountains. They are much smaller and
most of them herb-like, but the resemblance to a yellow single rose,
and the absence of thorns are common to them all. We have many
wild roses in this same family, of the genus Rosa, that have plenty of
thorns and closely resemble the red single roses of the garden.

30
 Pea Family
Prairie Pea, Lathyrus stipulaceus, B. AND ST. J.

Flowers, more than ½ inch across, are shaped like a cultivated sweet
pea, with very showy red banner and paler lateral petals and keel.
Plants, about 6 inches high, grow in irregular mats. The leaves are
pinnate, formed by about 4 pairs of narrow linear leaflets. These and
the stems are gray-green and, in most plains specimens, covered
with rather silky down. Grows in sandy soil on plains. Blooms May-
June.
This, and the quite different looking plants shown on the next three
pages, give but a small sample of the pea family, which is one of the
largest and most important of the plant groups. More than 150
species in this one family are native to Colorado, and additional ones
have been introduced for ornament or food. They take every form
and size from the little flat mats of deer clover, shown on the
opposite page, to the rank growing clumps of sweet clover that
spread themselves along our roads. Beans and alfalfa as well as
sweet peas, lupines and even locust trees, all belong to this big
family.

31
 Pea Family
Deer Clover, Trifolium nanum, TORR.

Individual flowers, pink-lavender to purple, formed along a keel, like

those of the cultivated clovers, about ½ inch long and rather slender,
grow singly or in twos or threes on short pedicels rising directly from
the root crown. The plant is a dense mat, often a foot or more
across, covered with small 3-foliate leaves. Grows on rocky flats or
slopes in alpine zone. Blooms June-July.
For many, acquainted only with the cultivated clovers of lawn and
meadow, it is a pleasure to know that the high pastures grazed by
deer and elk have clovers as well. At least three species are familiar
to observing travelers along Trail Ridge, or up Mt. Evans, or along any
road that crosses the enchanted land where trees stop and dwarfed
plant life takes over. The deer clover pictured here likes rocky places.
Its flowers are packed close together, but not clustered in heads as
are those of its alpine neighbor, Trifolium dasyphyllum, which closely
resembles the white clover of our lawns, though with touches on its
petals of red-brown. In the high places, extending down through the
sub-alpine zone there is also a bright red clover, Trifolium parryi,
smaller but otherwise much like the cultivated red clover.

32
 Pea Family
Lambert’s Loco, Oxytropis lambertii, PURSH
Individual flowers, about ½ inch wide, are formed of 5 dissimilar
petals, usually magenta red, sometimes other shades from rose to
purple. The banner bends back slightly and carries markings of
lighter color near its base; the 2 lateral petals are plain and angle
forward; the 2 lower petals form a narrow keel. Numerous flowers,
attached at the calyx base along the upper third of a naked stem,
form a showy spike 10 inches or more tall, several of which rise from
one root crown. Leaves, pinnate, with numerous green leaflets, rise
also from the root crown and are about half the height of the flower
spikes. Grows in foothills and higher parts of plains zone. Blooms
May-July.

The many members of the pea family going by the names of loco,
vetch, milk vetch, etc., are usually considered crass weeds and are in
disrepute because some of them are poisonous to stock. They often
grow in soil containing traces of selenium, and are doubly harmful in
that case. Where other browze is good, animals usually leave the
toxic ones alone, except the occasional horse that becomes “an
addict” and is “locoed.” In spite of these obnoxious qualities, there
are few plants that give more bright and decorative touches to the
plains.

33
Welcome to our website – the ideal destination for book lovers and
knowledge seekers. With a mission to inspire endlessly, we offer a
vast collection of books, ranging from classic literary works to
specialized publications, self-development books, and children's
literature. Each book is a new journey of discovery, expanding
knowledge and enriching the soul of the reade

Our website is not just a platform for buying books, but a bridge
connecting readers to the timeless values of culture and wisdom. With
an elegant, user-friendly interface and an intelligent search system,
we are committed to providing a quick and convenient shopping
experience. Additionally, our special promotions and home delivery
services ensure that you save time and fully enjoy the joy of reading.

Let us accompany you on the journey of exploring knowledge and

personal growth!

ebookball.com