Download the full version of the ebook at

https://ebookultra.com

Digital Image Sequence Processing Compression

and Analysis 1st Edition Todd R. Reed

https://ebookultra.com/download/digital-image-
sequence-processing-compression-and-analysis-1st-
edition-todd-r-reed/

Explore and download more ebook at https://ebookultra.com

 Recommended digital products (PDF, EPUB, MOBI) that
you can download immediately if you are interested.

Data Analysis Digital Signal and Image Processing 1st

Edition Gérard Govaert

https://ebookultra.com/download/data-analysis-digital-signal-and-
image-processing-1st-edition-gerard-govaert/

ebookultra.com

Digital Signal Processing Techniques and Applications in

Radar Image Processing 1st Edition Bu-Chin Wang

https://ebookultra.com/download/digital-signal-processing-techniques-
and-applications-in-radar-image-processing-1st-edition-bu-chin-wang/

ebookultra.com

Digital Image Processing PIKS Inside 3rd Edition William

Pratt

https://ebookultra.com/download/digital-image-processing-piks-
inside-3rd-edition-william-pratt/

ebookultra.com

Digital Image Processing PIKS Inside 3rd Edition William

K. Pratt

https://ebookultra.com/download/digital-image-processing-piks-
inside-3rd-edition-william-k-pratt/

ebookultra.com
Image and Video Compression and Multimedia 1st Edition
Sergios Theodoridis

https://ebookultra.com/download/image-and-video-compression-and-
multimedia-1st-edition-sergios-theodoridis/

ebookultra.com

Image sensors and signal processing for digital still

cameras 1st Edition Junichi Nakamura

https://ebookultra.com/download/image-sensors-and-signal-processing-
for-digital-still-cameras-1st-edition-junichi-nakamura/

ebookultra.com

Visual Perception Through Video Imagery Digital Signal and

Image Processing 1st Edition Michel Dhome

https://ebookultra.com/download/visual-perception-through-video-
imagery-digital-signal-and-image-processing-1st-edition-michel-dhome/

ebookultra.com

A Computational Introduction to Digital Image Processing

2nd Edition Edition Alasdair Mcandrew

https://ebookultra.com/download/a-computational-introduction-to-
digital-image-processing-2nd-edition-edition-alasdair-mcandrew/

ebookultra.com

Genetic and Evolutionary Computation for Image Processing

and Analysis 1st Edition Edited By: Stefano Cagnon

https://ebookultra.com/download/genetic-and-evolutionary-computation-
for-image-processing-and-analysis-1st-edition-edited-by-stefano-
cagnon/
ebookultra.com
Digital Image Sequence Processing Compression and
Analysis 1st Edition Todd R. Reed Digital Instant
Download
Author(s): Todd R. Reed
ISBN(s): 9780849315268, 0849315263
Edition: 1
File Details: PDF, 6.22 MB
Year: 2005
Language: english
 Digital Image
Sequence Processing,
Compression, and Analysis

© 2005 by CRC Press LLC


 Computer Engineering Series
Series Editor: Vojin Oklobdzija

Low-Power Electronics Design

Edited by Christian Piguet

Digital Image Sequence Processing,

Compression, and Analysis
Edited by Todd R. Reed

Coding and Signal Processing for

Magnetic Recording Systems
Edited by Bane Vasic and Erozan Kurtas

© 2005 by CRC Press LLC


 Digital Image
Sequence Processing,
Compression, and Analysis

EDITED BY

Todd R. Reed
University of Hawaii at Manoa
Honolulu, HI

CRC PR E S S
Boca Raton London New York Washington, D.C.
© 2005 by CRC Press LLC

 Library of Congress Cataloging-in-Publication Data

Digital image sequence processing, compression, and analysis / edited by Todd R. Reed.
p. cm.
Includes bibliographical references and index.
ISBN 0-8493-1526-3 (alk. paper)
1. Image processing—Digital techniques. 2. Digital video. I. Reed, Todd Randall.

TA1637.D536 2004
621.36e7—dc22 2004045491

This book contains information obtained from authentic and highly regarded sources. Reprinted material
is quoted with permission, and sources are indicated. A wide variety of references are listed. Reasonable
efforts have been made to publish reliable data and information, but the author and the publisher cannot
assume responsibility for the validity of all materials or for the consequences of their use.

Neither this book nor any part may be reproduced or transmitted in any form or by any means, electronic
or mechanical, including photocopying, microfilming, and recording, or by any information storage or
retrieval system, without prior permission in writing from the publisher.

All rights reserved. Authorization to photocopy items for internal or personal use, or the personal or
internal use of specific clients, may be granted by CRC Press LLC, provided that $1.50 per page
photocopied is paid directly to Copyright Clearance Center, 222 Rosewood Drive, Danvers, MA 01923
USA. The fee code for users of the Transactional Reporting Service is ISBN 0-8493-1526-
3/04/$0.00+$1.50. The fee is subject to change without notice. For organizations that have been granted
a photocopy license by the CCC, a separate system of payment has been arranged.

The consent of CRC Press LLC does not extend to copying for general distribution, for promotion, for
creating new works, or for resale. Specific permission must be obtained in writing from CRC Press LLC
for such copying.

Direct all inquiries to CRC Press LLC, 2000 N.W. Corporate Blvd., Boca Raton, Florida 33431.

Trademark Notice: Product or corporate names may be trademarks or registered trademarks, and are
used only for identification and explanation, without intent to infringe.

Visit the CRC Press Web site at www.crcpress.com

© 2005 by CRC Press LLC

No claim to original U.S. Government works

International Standard Book Number 0-8493-1526-3
Library of Congress Card Number 2004045491
Printed in the United States of America 1 2 3 4 5 6 7 8 9 0
Printed on acid-free paper

© 2005 by CRC Press LLC


 To my wife, Nancy.

© 2005 by CRC Press LLC


Preface
Digital image sequences (including digital video) are an increasingly com-
mon and important component in technical applications, ranging from med-
ical imaging and multimedia communications to autonomous vehicle navi-
gation. They are ubiquitous in the consumer domain, due to the immense
popularity of DVD video and the introduction of digital television.
Despite the fact that this form of visual representation has become com-
monplace, research involving digital image sequence remains extremely
active. The advent of increasingly economical sequence acquisition, storage,
and display devices, together with the widespread availability of inexpen-
sive computing power, opens new areas of investigation on an almost daily
basis.
The purpose of this work is to provide an overview of the current state
of the art, as viewed by the leading researchers in the field. In addition to
being an invaluable resource for those conducting or planning research in
this area, this book conveys a unified view of potential directions for indus-
trial development.

© 2005 by CRC Press LLC


About the Editor
Todd R. Reed received his B.S., M.S., and Ph.D. degrees in electrical
engineering from the University of Minnesota in 1977, 1986, and 1988,
respectively.
From 1977 to 1983, Dr. Reed worked as an electrical engineer at IBM
(San Jose, California; Rochester, Minnesota; and Boulder, Colorado) and from
1984 to 1986 he was a senior design engineer for Astrocom Corporation, St.
Paul, Minnesota. He served as a consultant to the MIT Lincoln Laboratory
from 1986 to 1988. In 1988, he was a visiting assistant professor in the
Department of Electrical Engineering, University of Minnesota. From 1989
to 1991, Dr. Reed acted as the head of the image sequence processing research
group in the Signal Processing Laboratory, Department of Electrical Engi-
neering, at the Swiss Federal Institute of Technology in Lausanne. From 1998
to 1999, he was a guest researcher in the Computer Vision Laboratory,
Department of Electrical Engineering, Linköping University, Sweden. From
2000 to 2002, he worked as an adjunct professor in the Programming Envi-
ronments Laboratory in the Department of Computer Science at Linköping.
From 1991 to 2002, he served on the faculty of the Department of Electrical
and Computer Engineering at the University of California, Davis. Dr. Reed
is currently professor and chair of the Department of Electrical Engineering
at the University of Hawaii, Manoa. His research interests include image
sequence processing and coding, multidimensional digital signal processing,
and computer vision.
Professor Reed is a senior member of the Institute of Electrical and
Electronics Engineers (IEEE) and a member of the European Association for
Signal Processing, the Association for Computing Machinery, the Society for
Industrial and Applied Mathematics, Tau Beta Pi, and Eta Kappa Nu.

© 2005 by CRC Press LLC


Contributors
Pedro M. Q. Aguiar Gaetano Giunta
ISR—Institute for Systems and Department of Applied Electronics
Robotics, IST—Instituto Superior University of Rome Tre
Técnico Rome, Italy
Lisboa, Portugal

Luis D. Alvarez Jan Horn

Department of Computer Science Institut für Mess- und
and A.I. Regelungstechnik
University of Granada Universität Karlsruhe
Granada, Spain Karlsruhe, Germany

Guido Maria Cortelazzo Radu S. Jasinschi

Department of Engineering Philips Research
Informatics Eindhoven, The Netherlands
University of Padova
Padova, Italy
Sören Kammel
Thao Dang Institut für Mess- und
Institut für Mess- und Regelungstechnik
Regelungstechnik Universität Karlsruhe
Universität Karlsruhe Karlsruhe, Germany
Karlsruhe, Germany
Aggelos K. Katsaggelos
Edward J. Delp Department of Electrical and
School of Electrical Engineering Computer Engineering
Purdue University Northwestern University
West Lafayette, Indiana, USA Evanston, Illinois, USA

Francesco G. B. De Natale Anil Kokaram

Dipartimento Informatica e Department of Electronic and
Telecomunicazioni Electrical Engineering
Universita di Trento University of Dublin
Trento, Italy Dublin, Ireland

© 2005 by CRC Press LLC


Luca Lucchese Charnchai Pluempitiwiriyawej
School of Engineering and Department of Electrical and
Computer Science Computer Engineering
Oregon State University Carnegie Mellon University
Corvallis, Oregon, USA Pittsburgh, Pennsylvania, USA

Rafael Molina
Christoph Stiller
Department of Computer Science
Institut für Mess- und
and A.I.
Regelungstechnik
University of Granada
Universität Karlsruhe
Granada, Spain
Karlsruhe, Germany
José M. F. Moura
Department of Electrical and Cuneyt M. Taskiran
Computer Engineering School of Electrical Engineering
Carnegie Mellon University Purdue University
Pittsburgh, Pennsylvania, USA West Lafayette, Indiana, USA

© 2005 by CRC Press LLC


Contents
Chapter 1 Introduction
Todd R. Reed

Chapter 2 Content-based image sequence representation

Pedro M. Q. Aguiar, Radu S. Jasinschi, José M. F. Moura, and
Charnchai Pluempitiwiriyawej

Chapter 3 The computation of motion

Christoph Stiller, Sören Kammel, Jan Horn, and Thao Dang

Chapter 4 Motion analysis and displacement estimation in the

frequency domain
Luca Lucchese and Guido Maria Cortelazzo

Chapter 5 Quality of service assessment in new generation

wireless video communications
Gaetano Giunta

Chapter 6 Error concealment in digital video

Francesco G. B. De Natale

Chapter 7 Image sequence restoration: A wider perspective

Anil Kokaram

Chapter 8 Video summarization

Cuneyt M. Taskiran and Edward J. Delp

Chapter 9 High-resolution images from a sequence of

low-resolution observations
Luis D. Alvarez, Rafael Molina, and Aggelos K. Katsaggelos

© 2005 by CRC Press LLC


chapter 1

Introduction
Todd R. Reed

The use of image sequences to depict motion dates back nearly two centuries.
One of the earlier approaches to motion picture “display” was invented in
1834 by the mathematician William George Horner. Originally called the
Daedaleum (after Daedalus, who was supposed to have made figures of men
that seemed to move), it was later called the Zoetrope (literally “life turning”)
or the Wheel of Life. The Daedaleum works by presenting a series of images,
one at a time, through slits in a circular drum as the drum is rotated.
Although this device is very simple, it illustrates some important con-
cepts that also underlie modern image sequence displays:

1. The impression of motion is illusory. It is the result of a property of

the visual system referred to as persistence of vision. An image is
perceived to remain for a period of time after it has been removed
from view. This illusion is the basis of all motion picture displays.
2. When the drum is rotated slowly, the images appear (as they are) a
disjoint sequence of still images. As the speed of rotation increases
and the images are displayed at a higher rate, a point is reached at
which motion is perceived, even though the images appear to flicker.
3. Further increasing the speed of rotation, we reach a point at which
flicker is no longer perceived (referred to as the critical fusion fre-
quency).
4. Finally, the slits in the drum illustrate a vital aspect of this illusion.
In order to perceive motion from a sequence of images, the stimulus
the individual images represent must be removed for a period of
time between each presentation. If not, the sequence of images simply
merges into a blur. No motion is perceived.

The attempt to display image sequences substantially predates the ability

to acquire them photographically. The first attempt to acquire a sequence of
photographs from an object in motion is reputed to have been inspired by

© 2005 by CRC Press LLC


a wager of Leland Stanford circa 1872. The wager involved whether or not,
at any time in its gait, a trotting horse has all four feet off the ground.
The apparatus that eventually resulted, built on Stanford’s estate in Palo
Alto by Eadweard Muybridge, consisted of a linear array of cameras whose
shutters are tripped in sequence as the subject passes each camera. This
device was used in 1878 to capture the first photographically recorded
(unposed) sequence. This is also the earliest known example of image
sequence analysis.
Although effective, Muybridge’s apparatus was not very portable. The
first portable motion picture camera was designed by E. J. Marey in 1882.
His “photographic gun” used dry plate technology to capture a series of 12
images in 1 second on a single disk. In that same year, Marey modified
Muybridge’s multicamera approach to use a single camera, repeatedly
exposing a plate via a rotating disk shutter. This device was used for motion
studies, utilizing white markers attached to key locations on a subject’s
anatomy (the hands, joints, feet, etc.). This basic approach is widely used
today for motion capture in animation.
Although of substantial technical and scientific interest, motion pictures
had little commercial promise until the invention of film by Hannibal Good-
win in 1887, and in 1889 by Henry W. Reichenbach for Eastman. This flexible
transparent substrate provided both a convenient carrier for the photo-
graphic emulsion and a means for viewing (or projecting) the sequence. A
great deal of activity ensued, including work sponsored by Thomas Edison
and conducted by his assistant, W. K. L. Dickson.
By 1895, a camera/projector system embodying key aspects of current
film standards (35-mm width, 24-frame-per-second frame rate) was devel-
oped by Louis Lumiére. This device was named the Cinématographe (hence
the cinéma).
The standardization of analog video in the early 1950s (NTSC) and late
1960s (SECAM and PAL) made motion pictures ubiquitous, with televisions
appearing in virtually every home in developed countries. Although these
systems were used primarily for entertainment purposes, systems for tech-
nical applications such as motion analysis continued to be developed.
Although not commercially successful, early attempts at video communica-
tion systems (e.g., by AT&T) also appeared during this time.
The advent of digital video standards in the 1990s (H.261, MPEG, and
those that followed), together with extremely inexpensive computing and
display platforms, has resulted in explosive growth in conventional (enter-
tainment) applications, in video communications, and in evolving areas such
as video interpretation and understanding.
In this book, we seek both to establish the current state of the art in the
utilization of digital image sequences and to indicate promising future direc-
tions for this field.
The choice of representation used in a video-processing, compression,
or analysis task is fundamental. The proper representation makes features
of interest apparent, significantly facilitating operations that follow. An inap-
© 2005 by CRC Press LLC

propriate representation obscures such features, adding significantly to com-
plexity (both conceptual and computational). In “Content-Based Image
Sequence Representation” by Aguiar, Jasinschi, Moura, and Pluempitiwir-
iyawej, video representations based on semantic content are examined. These
representations promise to be very powerful, enabling model-based and
object-based techniques in numerous applications. Examples include video
compression, video editing, video indexing, and scene understanding.
Motion analysis has been a primary motivation from the earliest days
of image sequence acquisition. More than 125 years later, the development
of motion analysis techniques remains a vibrant research area. Numerous
schools of thought can be identified. One useful classification is based on
the domain in which the analysis is conducted.
In “The Computation of Motion” by Stiller, Kammel, Horn, and Dang,
a survey and comparison of methods that could be classified as spatial
domain techniques are presented. These methods can be further categorized
as gradient-based, intensity-matching, and feature-matching algorithms. The
relative strengths of some of these approaches are illustrated in representa-
tive real-world applications.
An alternative class of motion analysis techniques has been developed
in the frequency (e.g., Fourier) domain. In addition to being analytically
intriguing, these methods correlate well with visual motion perception mod-
els. They also have practical advantages, such as robustness in the presence
of noise. In “Motion Analysis and Displacement Estimation in the Frequency
Domain” by Lucchese and Cortelazzo, methods of this type are examined
for planar rigid motion, planar affine motion, planar roto-translational dis-
placements, and planar affine displacements.
Although there remain technical issues surrounding wireless video com-
munications, economic considerations are of increasing importance. Quality
of service assurance is a critical component in the cost-effective deployment
of these systems. Customers should be guaranteed the quality of service for
which they pay. In “Quality of Service Assessment in New Generation Wire-
less Video Communications,” Giunta presents a discussion of quality-of-ser-
vice assessment methods for Third Generation (3G) wireless video commu-
nications. A novel technique based on embedded video watermarks is
introduced.
Wireless communications channels are extremely error-prone. While
error-correcting codes can be used, they impose computational overhead on
the sender and receiver and introduce redundancy into the transmitted
bitstream. However, in applications such as consumer-grade video commu-
nications, error-free transmission of all video data may be unnecessary if the
errors can be made unobtrusive. “Error Concealment in Digital Video” by
De Natale provides a survey and critical analysis of current techniques for
obscuring transmission errors in digital video.
With the increase in applications for digital media, the demand for con-
tent far exceeds production capabilities. This makes archived material, par-
ticularly motion picture film archives, increasingly valuable. Unfortunately,
© 2005 by CRC Press LLC

film is a very unstable means of archiving images, subject to a variety of
modes of degradation. The artifacts encountered in archived film, and algo-
rithms for correcting these artifacts, are discussed in “Image Sequence Res-
toration: A Wider Perspective” by Kokaram.
As digital video archives continue to grow, accessing these archives in
an efficient manner has become a critical issue. Concise condensations of
video material provide an effective means for browsing archives and may
also be useful for promoting the use of particular material. Approaches to
generating concise representations of video are examined in “Video Sum-
marization” by Taskiran and Delp.
Technological developments in video display have advanced very rap-
idly, to the point that affordable high-definition displays are widely available.
High definition program material, although produced at a growing rate, has
not kept pace. Furthermore, archival video may be available only at a fixed
(relatively low) resolution. In the final chapter of this book, “High-Resolution
Images from a Sequence of Low-Resolution Observations,” Alvarez, Molina,
and Katsaggelos examine approaches to producing high-definition material
from a low-definition source.

Bibliography
Gerald Mast. A Short History of Movies. The Bobbs-Merrill Company, Inc., New York,
1971.
Kenneth Macgowan. Behind the Screen – The History and Techniques of the Motion Picture.
Delacorte Press, New York, 1965.
C.W. Ceram. Archaeology of the Cinema. Harcourt, Brace & World, Inc., New York, 1965.
John Wyver. The Moving Image – An International History of Film, Television, and Video.
BFI Publishing, London, 1989.

© 2005 by CRC Press LLC


chapter 2

Content-based image
sequence representation
Pedro M. Q. Aguiar, Radu S. Jasinschi, José M. F. Moura, and
Charnchai Pluempitiwiriyawej1

Contents
2.1 Introduction
2.1.1 Mosaics for static 3-D scenes and large depth:
single layer
2.1.2 Mosaics for static 3-D scenes and variable depth:
multiple layers
2.1.3 Video representations with fully 3-D models
2.1.3.1 Structure from motion: factorization
2.2 Image segmentation
2.2.1 Calculus of variations
2.2.1.1 Adding constraints
2.2.1.2 Gradient descent flow
2.2.2 Overview of image segmentation methods
2.2.2.1 Edge-based approach
2.2.2.2 Region-based approach
2.2.3 Active contour methods
2.2.4 Parametric active contour
2.2.4.1 Variations of classical snakes
2.2.5 Curve evolution theory
2.2.6 Level set method
2.2.7 Geometric active contours

1 The work of the first author was partially supported by the (Portuguese) Foundation for

Science and Technology grant POSI/SRI/41561/2001. The work of the third and fourth authors
was partially supported by ONR grant # N000 14-00-1-0593 and by NIH grants R01EB/AI-00318
and P41EB001977.

© 2005 by CRC Press LLC


 2.2.8 STACS: Stochastic active contour scheme
2.3 Mosaics: From 2-D to 3-D
2.3.1 Generative video
2.3.1.1 Figure and background mosaics generation
2.3.2 3-D Based mosaics
2.3.2.1 Structure from motion: generalized eight-point
algorithm
2.3.2.2 Layered mosaics based on 3-D information
2.3.2.3 3-D mosaics
2.3.2.4 Summary
2.4 Three-dimensional object-based representation
2.4.1 3-D object modeling from video
2.4.1.1 Surface-based rank 1 factorization method
2.4.2 Framework
2.4.2.1 Image sequence representation
2.4.2.2 3-D motion representation
2.4.2.3 3-D shape representation
2.4.3 Video analysis
2.4.3.1 Image motion
2.4.3.2 3-D structure from 2-D motion
2.4.3.3 Translation estimation
2.4.3.4 Matrix of 2-D motion parameters
2.4.3.5 Rank 1 factorization
2.4.3.6 Decomposition stage
2.4.3.7 Normalization stage
2.4.3.8 Texture recovery
2.4.4 Video synthesis
2.4.5 Experiment
2.4.6 Applications
2.4.6.1 Video coding
2.4.6.2 Video content addressing
2.4.6.3 Virtualized reality
2.4.7 Summary
2.5 Conclusion
References

Abstract. In this chapter we overview methods that represent

video sequences in terms of their content. These methods differ
from those developed for MPEG/H.26X coding standards in that
sequences are described in terms of extended images instead of
collections of frames. We describe how these extended images,
e.g., mosaics, are generated by basically the same principle: the
incremental composition of visual photometric, geometric, and
multiview information into one or more extended images. Differ-
ent outputs, e.g., from single 2-D mosaics to full 3-D mosaics, are

© 2005 by CRC Press LLC


 obtained depending on the quality and quantity of photometric,
geometric, and multiview information. In particular, we detail a
framework well suited to the representation of scenes with inde-
pendently moving objects. We address the two following impor-
tant cases: (i) the moving objects can be represented by 2-D
silhouettes (generative video approach) or (ii) the camera motion
is such that the moving objects must be described by their 3-D
shape (recovered through rank 1 surface-based factorization). A
basic preprocessing step in content-based image sequence repre-
sentation is to extract and track the relevant background and
foreground objects. This is achieved by 2-D shape segmentation
for which there is a wealth of methods and approaches. The
chapter includes a brief description of active contour methods for
image segmentation.

2.1 Introduction
The processing, storage, and transmission of video sequences are now com-
mon features of many commercial and free products. In spite of the many
advances in the representation of video sequences, especially with the advent
and the development of the MPEG/H.26X video coding standards, there is
still room for more compact video representations than currently used by
these standards.
In this chapter we describe work developed in the last 20 years that
addresses the problem of content-based video representation. This work can
be seen as an evolution from standard computer vision, image processing,
computer graphics, and coding theory toward a full 3-D representation of
visual information. Major application domains using video sequences infor-
mation include visually guided robotics, inspection, and surveillance; and
visual rendering. In visually guided robotics, partial or full 3-D scene infor-
mation is necessary, which requires the full reconstruction of 3-D informa-
tion. On the other hand, inspection and surveillance robotics often requires
only 2-D information. In visual rendering, the main goal is to display the
video sequence in some device in the best visual quality manner. Common
to all these applications is the issue of compact representation since full quality
video requires an enormous amount of data, which makes its storage, pro-
cessing, and transmission a difficult problem. We consider in this paper a
hierarchy of content-based approaches: (i) generative video (GV) that gen-
eralizes 2-D mosaics; (ii) multilayered GV type representations; and (iii) full
3-D representation of objects.
The MPEG/H.26X standards use frame-based information. Frames are
represented by their GOP structure (e.g., IPPPBPPPBPPPBPPP), and each
frame is given by slices composed of macro-blocks that are made of typically
8 × 8 DCT blocks. In spite of many advances allowed by this representation,
it falls short in terms of the level of details represented and compression

© 2005 by CRC Press LLC


rates. DCT blocks for spatial luminance/color coding and macro-blocks for
motion coding provide the highest levels of details. However, they miss
capturing pixel-level luminance/color/texture spatial variations and tem-
poral (velocity) variations, thus leading to visual artifacts. The compression
ratios achieved, e.g., 40:1, are still too low for effective use of MPEG/H.26X
standards in multimedia applications for storage and communication pur-
poses.
Content-based representations go beyond frame-based or pixel-based
representations of sequences. Video content information is represented by
objects that have to be segmented and represented. These objects can be
based on 2-D information (e.g., faces, cars, or trees) or 3-D information (e.g.,
when faces, cars, or trees are represented in terms of their volumetric con-
tent). Just segmenting objects from individual video frames is not sufficient;
these segmented objects have to be combined across the sequence to generate
extended images for the same object. These extended images, which include
mosaics, are an important element in the “next generation” systems for
compact video representation. Extended images stand midway between
frame-based video representations and full 3-D representations. With
extended images, a more compact representation of videos is possible, which
allows for their more efficient processing, storage, and transmission.
In this chapter we discuss work on extended images as a sequence of
approaches that start with standard 2-D panoramas or mosaics, e.g., those
used in astronomy for very far objects, to full 3-D mosaics used in visually
guided robotics and augmented environments. In the evolution from stan-
dard single 2-D mosaics to full 3-D mosaics, more assumptions and infor-
mation about the 3-D world are used. We present this historical and tech-
nical evolution as the development of the same basic concept, i.e., the
incremental composition of photometric (luminance/color), shape (depth),
and points of view (multiview) information from successive frames in a
video sequence to generate one or more mosaics. As we make use of
additional assumptions and information about the world, we obtain dif-
ferent types of extended images.
One such content-based video representation is called generative video
(GV). In this representation, 2-D objects are segmented and compactly rep-
resented as, for example, coherent stacks of rectangles. These objects are then
used to generate mosaics. GV mosaics are different from standard mosaics.
GV mosaics include the static or slowly changing background mosaics, but
they also include foreground moving objects, which we call figures. The GV
video representation includes the following constructs: (i) layered mosaics,
one for each foreground moving 2-D object or objects lying at the same depth
level; and (ii) a set of operators that allow for the efficient synthesis of video
sequences. Depending on the relative depth between different objects in the
scene and the background, a single or a multilayered representation may be
needed. We have shown that GV allows for a very compact video sequence
representation, which enables a very efficient coding of videos with com-
pression ratios in the range of 1000:1.
© 2005 by CRC Press LLC

 Often, layered representations are not sufficient to describe well the
video sequence, for example, when the camera motion is such that the
rigidity of the real-world objects can only be captured by going beyond 2-D
shape models and resorting to fully 3-D models to describe the shape of the
objects. To recover automatically the 3-D shape of the objects and the 3-D
motion of the camera from the 2-D motion of the brightness pattern on the
image plane, we describe in this chapter the surface-based rank 1 factoriza-
tion method.
Content-based video representations, either single-layer or multi-
ple-layer GV, or full 3-D object representations involve as an important
preprocessing step the segmentation and tracking of 2-D objects. Segmenta-
tion is a very difficult problem for which there is a wealth of approaches
described in the literature. We discuss in this chapter contour-based methods
that are becoming popular. These methods are based on energy minimization
approaches and extend beyond the well-known “snakes” method in which
a set of points representing positions on the image boundary of 2-D objects
— contours — is tracked in time. These methods make certain assumptions
regarding the smoothness of these contours and how they evolve over time.
These assumptions are at the heart of representing “active” contours. For
completeness, we briefly discuss active-contour-based segmentation meth-
ods in this chapter.
In the next three subsections, we briefly overview work by others on
single- and multilayered video representations and 3-D representations. Sec-
tion 2.2 overviews active-contour-based approaches to segmentation. We
then focus in Section 2.3 on generative video and its generalizations to
multilayered representations and in Section 2.4 on the rank 1 surface-based
3-D video representations. Section 2.5 concludes the chapter.

2.1.1 Mosaics for static 3-D scenes and large depth: single layer
Image mosaics have received considerable attention from the fields of astron-
omy, biology, aerial photogrammetry, and image stabilization to video com-
pression, visualization, and virtualized environments, among others. The
main assumption in these application domains is that the 3-D scene layout
is given by static regions shown very far away from the camera, that is, with
large average depth values with respect to (w.r.t.) to the camera (center).
Methods using this assumption will be discussed next.
Lippman [1] developed the idea of mosaics in the context of video
production. This reference deals mostly with generating panoramic images
describing static background regions. In this technique, panoramic images
are generated by accumulating and integrating local image intensity infor-
mation. Objects moving in the scene are averaged out; their shape and
position in the image are described as a “halo” region containing the back-
ground region; the position of the object in the sequence is reconstructed by
appropriately matching the background region in the halo to that of the
background region in the enlarged image. Lippman’s target application is

© 2005 by CRC Press LLC


high-definition television (HDTV) systems that require the presentation of
video at different aspect ratios compared to standard TV. Burt and Adelson
[2] describe a multiresolution technique for image mosaicing. Their aim is
to generate photomosaics for which the region of spatial transition between
different images (or image parts) is smooth in terms of its gray level or color
difference. They use for this purpose Laplacian pyramid structures to decom-
pose each image into their component pass-band images defined at different
spatial resolution levels. For each band, they generate a mosaic, and the final
mosaic is given by combining the mosaics at the different pass-bands. Their
target applications are satellite imagery and computer graphics.
Hansen [3] and collaborators at the David Sarnoff Laboratory have
developed techniques for generating mosaics in the framework of military
reconnaissance, surveillance, and target detection. Their motivation is
image stabilization for systems moving at high speeds and that use, among
other things, video information. The successive images of these video
sequences display little overlap, and they show, in general, a static 3-D
scene and in some cases a single moving (target) object. Image or camera
stabilization is extremely difficult under these circumstances. Hansen and
coworkers use a mosaic-based stabilization technique by which a given
image of the video sequence is registered to the mosaic built from preceding
images of the sequence instead of just from the immediately preceding
image. This mosaic is called the reference mosaic. It describes an extended
view of a static 3-D terrain. The sequential mosaic generation is realized
through a series of image alignment operations, which include the estima-
tion of global image velocity and of image warping.
Teodosio and Bender [4] have proposed salient video stills as a novel
way to represent videos. A salient still represents the video sequence by a
single high-resolution image by translating, scaling, and warping images of
the sequence into a single high-resolution raster. This is realized by (i) cal-
culating the optical flow between successive images; (ii) using an affine
representation of the optical flow to appropriately translate, scale, and warp
images; and (iii) using a weighted median of the high-resolution image. As
an intermediate step, a continuous space–time raster is generated in order
to appropriately align all pixels, regardless of whether the camera pans or
zooms, thus creating the salient still.
Irani et al. [5] propose a video sequence representation in terms of static,
dynamic, and multiresolution mosaics. A static mosaic is built from collec-
tions of “submosaics,” one for each scene subsequence, by aligning all of its
frames to a fixed coordinate system. This type of mosaic can handle cases
of static scenes, but it is not adequate for one having temporally varying
information. In the latter case, a dynamic mosaic is built from a collection
of evolving mosaics. Each of these temporarily updated mosaics is updated
according to information from the most recent frame. One difference with
static mosaic generation is that the coordinate system of the dynamic mosaics
can be moving with the current frame. This allows for an efficient updating
of the dynamic content.
© 2005 by CRC Press LLC

2.1.2 Mosaics for static 3-D scenes and variable depth:
multiple layers
When a camera moves in a static scene containing fixed regions or objects
that cluster at different depth levels, it is necessary to generate multiple
mosaics, one for each layer.
Wang and Adelson [6] describe a method to generate layers of panoramic
images from video sequences generated through camera translation with
respect to static scenes. They use the information from the induced (camera)
motion. They segment the panoramic images into layers according to the
motion induced by the camera motion. Video mosaicing is pixel based. It
generates panoramic images from static scenery panned or zoomed by a
moving camera.

2.1.3 Video representations with fully 3-D models

The mosaicing approaches outlined above represent a video sequence in
terms of flat scenarios. Since the planar mosaics do not model the 3-D shape
of the objects, these approaches do not provide a clear separation among
object shape, motion, and texture. Although several researchers proposed
enhancing the mosaics by incorporating depth information (see, for example,
the plane + parallax approach [5, 7]), these models often do not provide
meaningful representations for the 3-D shape of the objects. In fact, any video
sequence obtained by rotating the camera around an object demands a
content-based representation that must be fully 3-D based.
Among 3-D-model-based video representations, the semantic coding
approach assumes that detailed a priori knowledge about the scene is avail-
able. An example of semantic coding is the utilization of head-and-shoulders
parametric models to represent facial video sequences (see [8, 9]). The video
analysis task estimates along time the small changes of the head-and-shoul-
ders model parameters. The video sequence is represented by the sequence
of estimated head-and-shoulders model parameters. This type of represen-
tation enables very high compression rates for the facial video sequences
but cannot cope with more general videos.
The use of 3-D-based representations for videos of general scenes
demands the automatic 3-D modeling of the environment. The information
source for a number of successful approaches to 3-D modeling has been a
range image (see, for example, [10, 11]).
This image, obtained from a range sensor, provides the depth between
the sensor and the environment facing it on a discrete grid. Since the range
image itself contains explicit information about the 3-D structure of the
environment, the references cited above deal with the problem of how to
combine a number of sets of 3-D points (each set corresponding to a range
image) into a 3-D model.
When no explicit 3-D information is given, the problem of computing
automatically a 3-D-model-based representation is that of building the 3-D

© 2005 by CRC Press LLC


models from the 2-D video data. The recovery of the 3-D structure (3-D shape
and 3-D motion) of rigid objects from 2-D video sequences has been widely
considered by the computer vision community. Methods that infer 3-D shape
from a single frame are based on cues such as shading and defocus. These
methods fail to give reliable 3-D shape estimates for unconstrained
real-world scenes. If no prior knowledge about the scene is available, the
cue to estimating the 3-D structure is the 2-D motion of the brightness pattern
in the image plane. For this reason, the problem is generally referred to as
structure from motion (SFM).

2.1.3.1 Structure from motion: factorization

Among the existing approaches to the multiframe SFM problem, the factoriza-
tion method [12] is an elegant method to recover structure from motion without
computing the absolute depth as an intermediate step. The object shape is
represented by the 3-D position of a set of feature points. The 2-D projection
of each feature point is tracked along the image sequence. The 3-D shape and
motion are then estimated by factorizing a measurement matrix whose columns
are the 2-D trajectories of each of the feature point projections. The factorization
method proved to be effective when processing videos obtained in controlled
environments with a relatively small number of feature points. However, to
provide dense depth estimates and dense descriptions of the shape, this method
usually requires hundreds of features, a situation that then poses a major
challenge in tracking these features along the image sequence and that leads
to a combinatorially complex correspondence problem.
In Section 2.4, we describe a 3-D-model-based video representation
scheme that overcomes this problem by using the surface-based rank 1
factorization method [13, 14]. There are two distinguishing features of this
approach. First, it is surface based rather than feature (point) based; i.e., it
describes the shape of the object by patches, e.g., planar patches or
higher-order polynomial patches. Planar patches provide not only localiza-
tion but also information regarding the orientation of the surface. To obtain
similar quality descriptions of the object, the number of patches needed is
usually much smaller than the number of feature points needed. In [13], it
is shown that the polynomial description of the patches leads to a parame-
terization of the object surface and this parametric description of the 3-D
shape induces a parametric model for the 2-D motion of the brightness
pattern in the image plane. Instead of tracking pointwise features, this
method tracks regions of many pixels, where the 2-D image motion of each
region is described by a single set of parameters. This approach avoids the
correspondence problem and is particularly suited for practical scenarios in
which the objects are, for example, large buildings that are well described
by piecewise flat surfaces. The second characteristic of the method in [13,
14] and in Section 2.4 is that it requires only the factorization of a rank 1
rather than rank 3 matrix, which simplifies significantly the computational
effort of the approach and is more robust to noise.

© 2005 by CRC Press LLC


 Clearly, the generation of images from 3-D models of the world is a subject
that has been addressed by the computer graphics community. When the
world models are inferred from photograph or video images, rather than
specified by an operator, the view generation process is known as image-based
rendering (IBR). Some systems use a set of calibrated cameras (i.e., with known
3-D positions and internal parameters) to capture the 3-D shape of the scene
and synthesize arbitrary views by texture mapping, e.g., the Virtualized Real-
ity system [15]. Other systems are tailored to the modeling of specific 3-D
objects like the Façade system [16], which does not need a priori calibration
but requires user interaction to establish point correspondences. These sys-
tems, as well as the framework described in Section 2.4, represent a scene by
using geometric models of the 3-D objects. A distinct approach to IBR uses
the plenoptic function [17] — an array that contains the light intensity as a
function of the viewing point position in 3-D space, the direction of propaga-
tion, the time, and the wavelength. If in empty space, the dependence on the
viewing point position along the direction of propagation may be dropped.
By dropping also the dependence on time, which assumes that the lighting
conditions are fixed, researchers have attempted to infer from images what
has been called the light field [18]. A major problem in rendering images from
acquired light fields is that, due to limitations on the number of images avail-
able and on the processing time, they are usually subsampled. The Lumigraph
system [19] overcomes this limitation by using the approximate geometry of
the 3-D scene to aid the interpolation of the light field.

2.2 Image segmentation

In this section, we discuss segmentation algorithms, in particular, energy
minimization and active-contour-based approaches, which are popularly
used in video image processing. In Subsection 2.2.1, we review concepts
from variational calculus and present several forms of the Euler-Lagrange
equation. In Subsection 2.2.2, we broadly classify the image segmentation
algorithms into two categories: edge-based and region-based. In Subsection
2.2.3, we consider active contour methods for image segmentation and
discuss their advantages and disadvantages. The seminal work on active
contours by Kass, Witkin, and Terzopoulos [20], including its variations,
is then discussed in Subsection 2.2.4. Next, we provide in Subsection 2.2.5
background on curve evolution, while Subsection 2.2.6 shows how curve
evolution can be implemented using the level set method. Finally, we
provide in Subsection 2.2.7 examples of segmentation by these geometric
active contour methods utilizing curve evolution theory and implemented
by the level set method.

2.2.1 Calculus of variations

In this subsection, we sketch the key concepts we need from the calculus of
variations, which are essential in the energy minimization approach to image
© 2005 by CRC Press LLC

processing. We present the Euler-Lagrange equation, provide a generic solution
when a constraint is added, and, finally, discuss gradient descent numerical
solutions.
Given a scalar function u( x):[0,1] q R with given constant boundary con-
ditions u(0)=a and u(1)=b, the basic problem in the calculus of variations is
to minimize an energy functional [21]

1
J (u) =
µ E(u(x), ue(x)) dx,
0
(2.1)

where E(u,ue) is a function of u and ue, the first derivative of u. From classical
calculus, we know that the extrema of a function f(x) in the interior of the
domain are attained at the zeros of the first derivative of f(x), i.e., where fe(x)
= 0. Similarly, to find the extrema of the functional J(u), we solve for the
zeros of the first variation of J, i.e., IJ = 0 . Let Iu and Iue be small perturba-
tions of u and ue, respectively. By Taylor series expansion of the integrand
in Equation (2.1), we have

yE yE
E(u + Iu, ue + Iue) = E(u, ue) + Iu + Iue + $. (2.2)
yu yu'

Then

1
J (u + Iu) = J (u) +
µ (E Iu + E
0
u ue Iue) dx + $ , (2.3)

yE yE
where Eu = and Eue = represent the partial derivatives of E(u,ue) with
yu yue
respect to u and ue, respectively. The first variation of J is then

IJ (u) = J (u + Iu)  J (u) (2.4)

=
µ (E Iu + E Iue)dx.
0
u ue (2.5)

Integrating by parts the second term of the integral, we have

1 1

µ Iu dx (E ) dx
d
Iue dx = Eue Iu( x) xx ==10 
µE
0
ue
0
ue (2.6)

µ Iu dx (E ) dx.
d
= ue (2.7)
0

© 2005 by CRC Press LLC


The nonintegral term vanishes because Iu(1) = Iu(0) = 0 due to the assumed
constant boundary conditions of u. Substituting Equation (2.7) back into
Equation (2.4), we obtain

µ ­®IuE  Iu dx (E )½¾ dx.

¬ d ¼
IJ (u) = u ue (2.8)
0

A necessary condition for u to be an extremum of J(u) is that u makes the

integrand zero, i.e.,

d yE d © yE ¹
Eu  E =  ª º = 0. (2.9)
dx ue yu dx « yue »

This is the Euler-Lagrange equation for a one-dimensional (1-D) problem in

the calculus of variations [21].
More generally, the Euler-Lagrange equation for an energy functional of
the form

1
J (u) =
µ E(x, u, ue, uee,$, u ) dx,
0
n
(2.10)

where un is the nth derivative of u(x) with respect to x, can be derived in a

similar manner as

d d2 dn
Eu  Eue + 2 Euee  ,$ , + (1)n n E n = 0. (2.11)
dx dx dx u

For a scalar function defined on a 2-D domain or a 2-D plane,

u( x , y ): R 2 q R , we have a similar result. For instance, given an energy
functional

J (u) =
µ µ E(u, u , u , u
<
x y xx , uyy ) dx dy , (2.12)

the corresponding Euler-Lagrange equation is given by

yE d © yE ¹ d © yE ¹ d 2 © yE ¹ d 2 © yE ¹
  ª º+ + ª º = 0. (2.13)
yu dx ª« yux º» dy « yuy » dx 2 ª« yuxx º» dy 2 « yuyy »

Analogously, we obtain a system of Euler-Lagrange equations for a

vector-value function u. For example, if u( x) = [u1( x) u2 ( x)]T : R q R 2 , then
the corresponding system of Euler-Lagrange equations is
© 2005 by CRC Press LLC

 d d2 dn
Eu  E e + 2 E ee  ,$ , + (1)n n E n = 0, (2.14)
1 dx 1 dx
u u1 dx u1

d d2 dn
Eu  E e + 2 E ee  ,$ , + (1)n n E n = 0. (2.15)
2 dx 2 dx
u u2 dx u2

2.2.1.1 Adding constraints

Usually, we are not allowed to freely search for the optimal u; rather, con-
straints are added. For instance, we may want to search for a function u(x)
that minimizes the energy functional

b
J1(u) =
µ E(x, u, ue) dx,
a
(2.16)

under a constraint functional

b
J 2 (u) =
µ G(x, u, ue) dx = c,
a
(2.17)

where c is a given constant. By use of a Lagrange multiplier Q , the new

energy functional becomes

J (u) = J1(u)  Q J 2 (u) (2.18)

b
=
µ [E(x, u, ue)  Q G(x, u, ue)] dx.
a
(2.19)

As a result, the corresponding Euler-Lagrange equation is

yE d © yG d ¹
 E Qª  G º = 0, (2.20)
yu dx ue « yu dx ue »

which must be solved subject to the constraint Equation (2.17).

2.2.1.2 Gradient descent flow

One of the fundamental questions in the calculus of variations is how to
solve the Euler-Lagrange equation, i.e., how to solve for u in

F (u) = 0, (2.21)

© 2005 by CRC Press LLC


where F (u) is a generic function of u whose zero makes the first variation
of a functional J zero, i.e., IJ = 0 . Equation (2.21) can be any of the
Euler-Lagrange equations in (2.11), (2.13), (2.14), or (2.20). Only in a very
limited number of simple cases is this problem solved analytically. In most
image-processing applications, directly solving this problem is infeasible.
One possible solution for F (u) = 0 is to first let u(x) be a function of an(other)
artificial time marching parameter t and then numerically solve the partial
differential equation (PDE)

yu
= F (u), (2.22)
yt

with a given initial u0(x) at t = 0. At steady state,

yu
=0 (2.23)
yt

implies that F (u) = 0 is achieved, and the solution to the Euler-Lagrange

equation is obtained. This is denoted as the gradient descent flow method.

2.2.2 Overview of image segmentation methods

Image segmentation is a fundamental step in building extended images, as
well as many other image- and video-processing techniques. The principal
goal of image segmentation is to partition an image into clusters or regions
that are homogeneous with respect to one or more characteristics or features.
The first major challenge in segmenting or partitioning an image is the
determination of the defining features that are unique to each meaningful
region so that they may be used to set that particular region apart from the
others. The defining features of each region manifest themselves in a variety
of ways, including, but not limited to, image intensity, color, surface lumi-
nance, and texture. In generative video and structure from motion, an impor-
tant feature is the 2-D-induced motion of the feature points or the surface
patches. Once the defining features are determined, the next challenging
problem is how to find the “best” way to capture these defining features
through some means such as statistical characteristics, transforms, decom-
positions, or other more complicated methodologies, and then use them to
partition the image efficiently. Furthermore, any corruption — by noise,
motion artifacts, and the missing data due to occlusion within the observed
image — poses additional problems to the segmentation process. Due to
these difficulties, the image segmentation problem remains a significant and
considerable challenge.
The image segmentation algorithms proposed thus far in the literature
may be broadly categorized into two different approaches, each with its own
strengths and weaknesses [22, 23]:

© 2005 by CRC Press LLC


 2.2.2.1 Edge-based approach
The edge-based approach relies on discontinuity in image features between
distinct regions. The goal of edge-based segmentation algorithms is to locate
the object boundaries, which separate distinct regions, at the points where the
image has high change (or gradient) in feature values. Most edge-based algo-
rithms exploit spatial information by examining local edges found within the
image. They are often very easy to implement and quick to compute, as they
involve a local convolution of the observed image with a gradient filter. More-
over, they do not require a priori information about image content. The Sobel
[24], Prewitt [25], Laplacian [26, 27], or Canny [28] edge detectors are just a
few examples. For simple noise-free images, detection of edges results in
straightforward boundary delineation. However, when applied to noisy or
complex images, edge detectors have three major problems:

1. They are very sensitive to noise.

2. They require a selection of an edge threshold.
3. They do not generate a complete boundary of the object because the
edges often do not enclose the object completely due to noise or
artifacts in the image or the touching or overlapping of objects.

These obstacles are difficult to overcome because solving one usually

leads to added problems in the others. To reduce the effect of the noise, one
may lowpass filter the image before applying an edge operator. However,
lowpass filtering also suppresses soft edges, which in turn leads to more
incomplete edges to distinguish the object boundary. On the other hand, to
obtain more complete edges, one may lower the threshold to be more sen-
sitive to, and thus include more, weak edges, but this means more spurious
edges appear due to noise. To obtain satisfactory segmentation results from
edge-based techniques, an ad hoc postprocessing method such as the vector
graph method of Casadei and Mitter [29, 30] is often required after the edge
detection to link or group edges that correspond to the same object boundary
and get rid of other spurious edges. However, such an automatic edge
linking algorithm is computationally expensive and generally not very reli-
able.

2.2.2.2 Region-based approach

The region-based approach, as opposed to the edge-based approach, relies
on the similarity of patterns in image features within a cluster of neighboring
pixels. Region-based techniques, such as region growing or region merging
[31, 32, 33], assign membership to objects based on homogeneous statistics.
The statistics are generated and updated dynamically. Region-growing meth-
ods generate a segmentation map by starting with small regions that belong
to the structure of interest, called seeds. To grow the seeds into larger regions,
the neighboring pixels are then examined one at a time. If they are sufficiently
similar to the seeds, based on a uniformity test, then they are assigned to

© 2005 by CRC Press LLC


the growing region. The procedure continues until no more pixels can be
added. The seeding scheme to create the initial regions and the homogeneity
criteria for when and how to merge regions are determined a priori. The
advantage of region-based models is that the statistics of the entire image,
rather than local image information, are considered. As a result, the tech-
niques are robust to noise and can be used to locate boundaries that do not
correspond to large image gradients. However, there is no provision in the
region-based framework to include the object boundary in the decision-mak-
ing process, which usually leads to irregular or noisy boundaries and holes
in the interior of the object. Moreover, the seeds have to be initially picked
(usually by an operator) to be within the region of interest, or else the result
may be undesirable.

2.2.3 Active contour methods

Among a wide variety of segmentation algorithms, active contour methods
[20, 34–42] have received considerable interest, particularly in the video
image–processing community. The first active contour method, called
“snake,” was introduced in 1987 by Kass, Witkin, and Terzopoulos [20, 34].
Since then the techniques of active contours for image segmentation have
grown significantly and have been used in other applications as well. An
extensive discussion of various segmentation methods as well as a large set
of references on the subject may be found in [43].
Because active contour methods deform a closed contour, this segmen-
tation technique guarantees continuous closed boundaries in the resulting
segmentation. In principle, active contour methods involve the evolution of
curves toward the boundary of an object through the solution of an energy
functional minimization problem. The energy functionals in active contour
models depend not only on the image properties but also on the shape of
the contour. Therefore, they are considered a high-level image segmentation
scheme, as opposed to the traditional low-level schemes such as edge detec-
tors [24, 28] or region-growing methods [31, 32, 33].
The evolution of the active contours is often described by a PDE, which
can be tracked either by a straightforward numerical scheme such as the
Lagrangian parameterized control points [44] or by more sophisticated
numerical schemes such as the Eulerian level set methods [45, 46].
Although traditional active contours for image segmentation are edge
based, the current trends are region-based active contours [40, 42] or hybrid
active contour models, which utilize both region-based and edge-based
information [39, 41]. This is because the region-based models, which rely on
regional statistics for segmentation, are more robust to noise and less sensi-
tive to the placement of the initial contour than the edge-based models.
The classical snake algorithm [20] works explicitly with a parameterized
curve. Thus, it is also referred to as a parametric active contour, in contrast
to the geometric active contour [47], which is based on the theory of curve

© 2005 by CRC Press LLC


evolution. Unlike the parametric active contour methods, the geometric
active contour methods are usually implemented implicitly through level
sets [45, 46].
In the following subsections, we describe the parametric active contour
method, or classical snakes, and discuss its advantages and its shortcomings
in Subsection 2.2.4. We also present two variations of classical snakes that
attempt to improve the snake algorithms. We then provide background on
the contour evolution theory and the level set method in Subsections 2.2.5
and 2.2.6, respectively. We finally show in Subsection 2.2.7 how the geometric
contour method, which is based on the curve evolution theory and often
implemented by the level set method, can improve the performance of image
segmentation over the parametric active contour-based algorithms.

2.2.4 Parametric active contour

The parametric active contour model or snake algorithm [20] was first intro-
duced in the computer vision community to overcome the traditional reli-
ance on low-level image features like pixel intensities. The active contour
model is considered a high-level mechanism because it imposes the shape
model of the object in the processing. The snake algorithm turns the bound-
ary extraction problem into an energy minimization problem [48]. A tradi-
tional snake is a parameterized curve C( p) = [ x( p) y( p) ]T for p ‘[0,1] that
moves through a spatial domain < of the image I(x,y) to minimize the energy
functional

J (C) = J int (C) + J ext (C). (2.24)

It has two energy components, the internal energy J int and the external
energy J ext . The high-level shape model of the object is controlled by the
internal energy, whereas the external energy is designed to capture the
low-level features of interest, very often edges. The main idea is to minimize
these two energies simultaneously. To control the smoothness and the con-
tinuity of the curve, the internal energy governs the first and second deriv-
atives of the contour, i.e.,

1
1 2 2
J int = µ F Ce(p) + G Cee( p) dp , (2.25)
2 0

where F and G are constants and Ce( p) and Cee( p) are the first and second
derivatives of the contour with respect to the indexing variable p, respec-
tively. The first derivative discourages stretching and makes the contour
behave like an elastic string. The second derivative discourages bending and
makes it behave like a rigid rod. Therefore, the weighting parameters F and
G are used to control the strength of the model’s elasticity and rigidity,
respectively.
© 2005 by CRC Press LLC

 The external energy, on the other hand, is computed by integrating a
potential energy function P(x,y) along the contour C( p), i.e.,

1
J ext = µ P(C(p)) dp, (2.26)
0

where P(x,y) is derived from the image data. The potential energy function
P(x,y) must take small values at the salient features of interest because the
contour C( p) is to search for the minimum external energy. Given a gray-level
image I(x,y), viewed as a function of the continuous variables (x,y), a typical
potential energy function designed for the active contour C that captures
step edges is

P( x , y ) = |”GX ( x , y ) * I ( x , y )|2 , (2.27)

where GX ( x , y ) is a 2-D Gaussian function with variance X 2 , ” represents

the gradient operator, and  is the image convolution operator. The potential
energy function defined as in (2.27) is called the edge map of the image. Figure
2.1(b) shows the corresponding edge map of the image in Figure 2.1(a).
The problem of finding a curve C(p) that minimizes an energy functional
J(C(p)) is known as a variational problem [21]. It has been shown in [20] that
the curve C that minimizes J(C) in (2.24) must satisfy the following
Euler-Lagrange equation

FCee( p)  GCeeee( p)  ”P(C( p)) = 0. (2.28)

To find a solution to Equation (2.28), the snake is made dynamic by first

letting the contour C(p) be a function of time t (as well as p), i.e., C(p,t), and
then replacing the 0 on the right-hand side of Equation (2.28) by the partial
derivative of C with respect to t as the following

yC
= FCee( p)  GCeeee( p)  ”P(C( p)). (2.29)
yt

The gradient descent method is then used to iteratively solve for the zero
of (2.29).
To gain some insight about the physical behavior of the evolution of
active contours, Xu and Prince realized Equation (2.29) as the balancing
between two forces [38]

yC
= Fint (C) + Fext (C), (2.30)
yt

where the internal force is given by

© 2005 by CRC Press LLC

 Fint = FCee( p)  GCeeee( p), (2.31)

and the external force is given by

Fext = ”P( x , y ). (2.32)

The internal force Fint dictates the regularity of the contour, whereas the
external force Fext pulls it toward the desired image feature. We call Fext the
potential force field, because it is the vector field that pulls the evolving contour
toward the desired feature (edges) in the image. Figure 2.1(c) shows the
potential force field, which is the negative gradient magnitude of the edge
map in Figure 2.1(b). Figure 2.1(d) zooms in the area within the square box
shown in Figure 2.1(c).
The snake algorithm gains its popularity in the computer vision com-
munity because of the following characteristics:

1. It is deformable, which means it can be applied to segment objects

with various shapes and sizes.
2. It guarantees a smooth and closed boundary of the object.
3. It has been proven very useful in motion tracking for video.

The major drawbacks associated with the snake’s edge-based approach are:

1. It is very sensitive to noise because it requires the use of differential

operators to calculate the edge map.
2. The potential forces in the potential force field are only present in
the close vicinity of high values in the edge map.
3. It utilizes only the local information along the object boundaries, not
the entire image.

Hence, for the snake algorithm to converge to a desirable result, the

initial contour must be placed close enough to the true boundary of the
object. Otherwise, the evolving contour might stop at undesirable spurious
edges or the contour might not move at all if the potential force on the
contour front is not present. As a result, the initial contour is often obtained
manually. This is a key pitfall of the snake method.

2.2.4.1 Variations of classical snakes

Many efforts have been made to address the limitations of the original snakes
method. For example, to help the snake move or avoid being trapped by
spurious isolated edge points, Cohen’s balloon snake approach [35] added
another artificial inflation force to the external force component of Equation
(2.30). Thus, the balloon snake’s external force becomes

© 2005 by CRC Press LLC


 a. b.

c. d.

Figure 2.1 (a) Original image; (b) edge map derived from the original image (a);
(c) potential force field: the negative gradient of the edge map (b); (d) zoom-in of
area within the square box in (c).

Fext = ”P( x , y ) + Fconst nˆ , (2.33)

where Fconst is an arbitrary constant and n̂ is the unit normal vector on the
contour front. However, the balloon snake has limitations. Although the
balloon snake aims to pass through edges that are too weak with respect to
the inflation force Fconst n̂ , adjusting the strength of the balloon force is diffi-
cult because it must be large enough to overcome the weak edges and noises
but small enough not to overwhelm a legitimate boundary. Besides, the
balloon force is image independent; i.e., it is not derived from the image.
Therefore, the contour will continue to inflate at the points where the true
boundary is missing or weaker than the inflation force.
Xu and Prince [38, 49] introduced a new external force for edge-based
snakes called the gradient vector flow (GVF) snake. In their method, instead

© 2005 by CRC Press LLC


of directly using the gradient of the edge map as the potential force field,
they diffuse it first to obtain a new force field that has a larger capture range
than the gradient of the edge map. Figures 2.2(a) and (b) depict the gradient
of an edge map and the Xu and Prince’s new force field, respectively. Com-
paring the two figures, we observe that the Xu and Prince’s vector forces
gradually decrease as they are away from the edge pixels, whereas the vector
forces in the gradient of the edge map exist only in the neighboring pixels
of the edge pixels. As a result, there are no forces to pull a contour that is
located at the pixels far away from the edge pixels in the gradient of the
edge map field, but the contour may experience some forces at the same
location in the Xu and Prince’s force field.
Two other limitations associated with the parametric representation of
the classical snake algorithm are the need to perform reparameterization
and topological adaptation. It is often necessary to dynamically reparam-
eterize the snake in order to maintain a faithful delineation of the object
boundary. This adds computational overhead to the algorithm. In addition,
when the contour fragments need to be merged or split, it may require a
new topology and, thus, the reconstruction of the new parameterization.
McInerney and Terzopoulos [50] have proposed an algorithm to address
this problem.

2.2.5 Curve evolution theory

In this subsection, we explain how to control the motion of a propagating
contour using the theory of curve evolution. In particular, we present two
examples of the motion for a propagating curve that are commonly used in
active contour schemes for image segmentation.
Denote a family of smooth contours as

¬x( p , t)¼
C( p , t) = ­ ½, (2.34)
® y( p , t)¾

where p ‘[0,1] parameterizes the set of points on each curve, and t ‘[0, h)
parameterizes the family of curves at different time evolutions. With this
parameterization scheme, a closed contour has the property that

C(0, t) = C(1, t)  t. (2.35)

We are interested in finding an equation for the propagating motion of a

curve that eventually segments the image. Assume a variational approach
for image segmentation formulated as finding the curve C such that

C = argmin J(C), (2.36)

C

© 2005 by CRC Press LLC


 a.

b.

Figure 2.2 Two examples of the potential force fields of an edge map: (a) gradient of
the edge map; (b) Xu and Prince’s GVF field.

where J is an energy functional constructed to capture all the criteria that

lead to the desired segmentation. The solution to this variational problem
often involves a PDE.
Let F(C) denote an Euler-Lagrange equation such that the first variation
of J(C) with respect to the contour C is zero. Under general assumptions,
the necessary condition for C to be the minimizer of J(C) is that F(C) = 0.
The solution to this necessary condition can be computed as the steady state
solution of the following PDE [51]

yC
= F(C). (2.37)
yt

This equation is the curve evolution equation or the flow for the curve C.
The form of this equation indicates that F(C) represents the “force” acting
upon the contour front. It can also be viewed as the velocity at which the

© 2005 by CRC Press LLC


contour evolves. Generally, the force F has two components. As depicted in
Figure 2.3, Fn̂ is the component of F that points in the normal direction with
respect to the contour front, and Ft̂ is the (other) component of F that is
tangent to C.
In curve evolution theory, we are interested only in Fn̂ because it is the
force that moves the contour front forward (or inward), hence changing the
geometry of the contour. The flow along Ft̂ , on the other hand, only repa-
rameterizes the curve and does not play any role in the evolution of the
curve. Therefore, the curve evolution equation is often reduced to just the
normal component as

yC
= F nˆ , (2.38)
yt

where F is called the speed function. In principle, the speed function depends
on the local and global properties of the contour. Local properties of the
contour include local geometric information such as the contour’s principal
curvature P or the unit normal vector n̂ of the contour. Global properties of
the curve depend on its shape and position.
Coming up with an appropriate speed function, or equivalently the
curve evolution equation, for the image segmentation underlies much of the
research in this field. As an example, consider the Euclidean curve-shorten-
ing flow given by

yC
= P nˆ . (2.39)
yt

This flow corresponds to the gradient descent along the direction in which
the Euclidean arc length of the curve

Fn

Ft

Figure 2.3 The normal and tangential components of a force on the contour front.

© 2005 by CRC Press LLC


 L=
µ ds
C
(2.40)

decreases most rapidly. As shown in Figure 2.4, a jagged closed contour

evolving under this flow becomes smoother. Flow (2.39) has a number of
attractive properties, which make it very useful in a range of image-process-
ing applications. However, it is never used alone because if we continue the
evolution with this flow, the curve will shrink to a circle, then to a point,
and then finally vanishes.
Another example illustrates some of the problems associated with a
propagating curve. Consider the curve-evolution equation

yC
= Vonˆ , (2.41)
yt

where Vo is a constant. If Vo is positive, the contour inflates. If Vo is negative,

the contour evolves in a deflationary fashion. This is because it corresponds
to the minimization of the area within the closed contour.
As seen in Figure 2.5, most curves evolving under the constant flow
(2.41) often develop sharp points or corners that are nondifferentiable (along
the contour). These singularities pose a problem of how to continue imple-
menting the next evolution of the curve because the normal to the curve at
a singular point is ambiguous. However, an elegant numerical implementa-
tion through the level set method provides an “entropy solution” that solves
this curve evolution problem [45, 46, 52, 53]. Malladi et al. [37] and Caselles
et al. [54] utilized both the curvature flow (2.39) and the constant flow (2.41)
in their active contour schemes for image segmentation because they are
complementary to each other. Whereas the constant flow can create singu-
larities from an initial smooth contour, the curvature flow removes them by
smoothing the contour in the process.

Figure 2.4 Flow under curvature: a jagged contour becomes smoother.

© 2005 by CRC Press LLC


Figure 2.5 Flow with negative constant speed deflates the contour.

2.2.6 Level set method

Given a current position for the contour C and the equation for its motion
such as the one in (2.37), we need a method to track this curve as it evolves.
In general, there are two approaches to track the contour, the Lagrangian
and the Eulerian approaches. The Lagrangian approach is a straightforward
difference approximation scheme. It parameterizes the contour discretely
into a set of control points lying along the moving front. The motion vectors,
derived from the curve-evolution equation through a difference approxima-
tion scheme, are then applied to these control points to move the contour
front. The control points then advance to their new locations to represent
the updated curve front. Though this is a natural approach to track the
evolving contour, the approach suffers from several problems [55]:

1. This approach requires an impractically small time step to achieve a

stable evolution.
2. As the curve evolves, the control points tend to “clump” together
near high curvature regions, causing numerical instability. Methods
for control points reparameterization are then needed, but they are
often less than perfect and hence can give rise to errors.
3. Besides numerical instability, there are also problems associated with
the way the Lagrangian approach handles topological changes. As
the curve splits or merges, topological problems occur, requiring ad
hoc techniques [50, 56] to continue to make this approach work.

Osher and Sethian [45, 46, 52, 53] developed the level set technique for
tracking curves in the Eulerian framework, written in terms of a fixed coor-
dinate system. There are four main advantages to this level set technique:

1. Since the underlying coordinate system is fixed, discrete mesh points

do not move; the instability problems of the Lagrangian approxima-
tions can be avoided.
2. Topological changes are handled naturally and automatically.
3. The moving front is accurately captured regardless of whether it
contains cusps or sharp corners.

© 2005 by CRC Press LLC


 4. The technique can be extended to work on any number of spatial
dimensions.

The level set method [45] implicitly represents the evolving contour C(t) by
embedding it as the zero level of a level set function K : R 2 × [0, h) q R , i.e.,

{ }
C(t) = ( x , y ) ‘ < : K ( x , y , t) = 0 . (2.42)

Starting with an initial level set function K (t = 0), we then evolve K (t) so that
its zero level set moves according to the desired flow of the contour. Based
on the convention that this level set graph has negative values inside C and
positive values outside C, i.e.,

{
inside(C) = <1 = ( x , y ) ‘ < : K ( x , y , t) > 0 ,} (2.43)

{ }
outside(C) = < 2 = ( x , y ) ‘ < : K ( x , y , t) < 0 , (2.44)

the level set function K can be implemented as the signed Euclidean distance
to the contour C. For details about how to implement the Euclidean distance
to a contour, see [46, 57]. Using the standard Heaviside function

¯ 1, if K v0
H (K ) = ° , (2.45)
± 0, if K <0

we can conveniently mask out the image pixels that are inside, outside, or on
the contour C. For instance, the function H (K ) represents the binary template
of the image pixels that are inside or on the contour. The function 1  H (K )
represents the binary template of the image pixels that are strictly outside the
contour. To select only the pixels that are on the contour C, we can
use H (K )  [1  H ( K )] . To facilitate numerical implementation, however, the
regularized Heaviside function and its derivative, the regularized delta func-
tion, are often used instead. Define the regularized Heaviside function by

1¬ 2 ©K¹¼
H ‘(K ) = ­1 + arctan (2.46)
2® U « ‘» ½¾

then the regularized delta function is

d
I ‘(K ) = H (K ), (2.47)
dK ‘

© 2005 by CRC Press LLC


 1¬ ‘ ¼
= . (2.48)
U ­® ‘2 +K 2 ½¾

The functions H ‘(K ) , 1  H ‘(K ) , and I ‘(K ) are to represent the templates of
the image pixels that are inside, outside, and on the contour C, respectively.
By defining the sign of the level set function K to be positive inside and
negative outside the contour, the unit normal vector n of the contour C,
defined as

”K
n= , (2.49)
”K

will point inward as shown in Figure 2.6. Furthermore, the curvature P along
the contour, defined as

© ”K ¹ KxxKy2  2KxKyKxy + KyyKx2

P = div (n) = div ª º= (2.50)
( )
3/2
« ”K » K2 + K2 x y

is positive where the unit normal vectors diverge. On the other hand, the
curvature of the contour is negative if the unit normal vectors converge (see
Figure 2.6).

2.2.7 Geometric active contours

Based on the theory of curve evolution [58], geometric active contours [37,
47] evolve the curves using only geometric measures, such as the curvature
and the normal vectors, resulting in a contour evolution that is independent
of the curve’s parameterization. Therefore, there is no need for reparameter-
ization. In addition, the geometric active contour method can be implicitly

C :φ = 0

φ>0
κ>0

κ<0
φ<0

Figure 2.6 Unit normal vectors and curvature of the contour C.

© 2005 by CRC Press LLC


implemented by the level set technique [46], which handles the topology
change automatically.
As mentioned before in Subsection 2.2.5, Malladi, Sethian, and Vemuri
[37] and Caselles, Kimmel, and Sapiro [54] utilized curvature flow (2.39) and
the constant flow (2.41) concurrently to move the contour C in the direction
of its normal vector n as in

yC
=  g š (P + Vo ) n , (2.51)
yt

where P is the curvature of the contour; Vo is a constant; and g, designed to

capture prominent edges, is a decreasing function of the gradient image (or
the edge map). An example of a suitable g is

1
g= . (2.52)
1 + ”G  I ( x , y)

We can see that the contour front will slow down where the value of the
edge map ”G  I ( x , y) is high because g approaches zero, but it will keep
moving at constant speed Vo where the edge map value is zero. Therefore,
the effect of the constant term Vo is the same as Cohen’s balloon force [35].
As mentioned before, the effect of the curvature term P is only to smooth
out the contour [45]. Hence, it plays the role of the internal energy term in
the classical snake method [20].
This scheme works well for objects that have well-defined edge maps.
However, when the object boundary is difficult to distinguish, the evolving
contour may leak out because the multiplicative term g only slows down
the contour near the edges. It does not completely stop it. Chan and Vese
[42] describe a new active contour scheme that does not use edges but
combines an energy minimization approach with a level set-based solution.
If C is a contour that partitions the domain of an image I(x,y) into two regions,
the region inside the contour <1 and the region outside the contour <2 , then
their approach is to minimize the functional

J (C) = µ š Length(C)
2 2 (2.53)
+ Q1 I ( x , y )  c1 dxdy + Q 2
µ
<1
µ
<2
I ( x , y )  c2 dxdy ,

where µ , Q 1 , and Q 2 are constants and c1 and c2 are the average intensity
values of the image pixels inside and outside the contour, respectively. As
defined in (2.43) and (2.44), <1 and <2 represent the pixels inside and out-
side the contour C, respectively. If C is embedded as a zero level of the level
set function K , minimizing the functional (2.53) is equivalent to solving the
PDE [42]
© 2005 by CRC Press LLC

 yK ¬ © ”K ¹ ¼
= ­ µ divª º  Q 1( I  c1 ) + Q 2 ( I  c2 ) ½I ‘(K ),
2 2
(2.54)
yt ® «|”K |» ¾

where div(š) denotes the divergence and I ‘(K ) is the regularized delta func-
tion defined in (2.47). It masks out only the zero level set of K , i.e., the contour
C. The first term, the divergence term, which only affects the smoothness of
the contour, is actually the motion under the curvature [45] because

© ”K ¹
div ª º I (K ) = P (C), (2.55)
«|”K |»

where P (C) denotes the curvature of the contour C; see Equation (2.50).
Therefore, Equation (2.54) is equivalent to the curve evolution

yC
yt
[ ]
= µ P  Q 1( I  c1 )2 + Q 2 ( I  c2 )2 n , (2.56)

where n, as defined in (2.49), is the outward unit normal vector of the contour
C. The last two terms of (2.54), however, work together to move the contour
such that all the pixels whose intensity values are similar to c1 are grouped
within the contour, and all the pixels whose intensity values are close to c2
are assigned to be outside the contour.
As a result, the image will be segmented into two constant intensity
regions. More importantly, unlike an edge-based model, this region-based
geometric active contour model is less sensitive to the initial location of the
contour. In addition, with the level set implementation, the algorithm han-
dles the topological changes of the contour automatically when contour
splitting or merging occurs. Figure 2.7(a) shows the process of segmenting
a blurred-edge, piecewise-constant image with a hole in it using Chan and
Vese’s method, and Figure 2.7(b) depicts the final result. The initial contour,
the dashed line, is simply an ellipse in the middle of the image and the final
contours are shown as the two solid lines in Figure 2.7(b); one is the outside
boundary of the object and the other is at the boundary of the hole within
the object. We can see that this successful result is achieved even when the
initial contour is far from the object’s true boundary. Moreover, the splitting
and merging of the contour around the hole, as seen in Figure 2.7(a), is done
automatically through the level set implementation.

2.2.8 STACS: Stochastic active contour scheme

The segmentation approaches discussed in the previous subsections often
work well but also fail in many important applications. For example, the
approach in [42] can segment reasonably well an object from the background

© 2005 by CRC Press LLC


 a.

b.

Figure 2.7 Image segmentation using Chan and Vese’s method: (a) evolving contour;
(b) final result.

when the pixels inside and outside the contour follow the two-value model
assumption well. In practice this is usually not the case, and this method
may lead to poor segmentation results. Another common insufficiency of
the segmentation methods presented in the previous subsections is that often
we have a good indication of the shape of the object but the segmentation
method has no explicit way to account for this knowledge. Edge- and
region-based information may both be important clues, but existing methods
usually take one or the other into account, not both.
We have developed in [59, 60] a method that we call the Stochastic Active
Contour Scheme (STACS), which addresses these insufficiencies of existing
approaches. It is an energy-based minimization approach in which the
energy functional combines four terms: (i) an edge-based term; (ii) a
region-based term that models the image textures stochastically rather than

© 2005 by CRC Press LLC


deterministically; (iii) a contour smoothness-based term; and, finally, (iv) a
shape prior term. These four terms in the energy functional lead to very
good segmentation results, even when the objects to be segmented exhibit
low contrast with respect to neighboring pixels or the texture of different
objects is quite similar. To enhance performance, STACS implements an
annealing schedule on the relative weights among the four energy terms.
This addresses the following issue. In practice, it is not clear which clue
should dominate the segmentation; it is commonly true that, at the beginning
of the segmentation process, edge- or region-based information is the more
important clue, but, as the contour evolves, shape may become the dominant
clue. STACS places initially more weight in the edge- and region-based terms
but then slowly adapts the weights to reverse this relative importance so
that, toward the end of the segmentation, more weight is placed in the shape
prior and contour smoothness constraint terms. Lack of space prevents us
from illustrating the good performance of STACS; details are in [59, 60] and
references therein.

2.3 Mosaics: From 2-D to 3-D

In this section we will describe two approaches to mosaic generation. The
first is called generative video [61, 62, 63, 64] and the second one [65, 66, 67]
uses partial or full 3-D information to generate mosaics. The motivation in
these two approaches is to overcome the limitations of classical mosaic
generation methods; the objects in the (3-D) scene are very far from the
camera so that there barely exists any parallax2. Depending on the relative
geometry between the camera and the scene, different methods can be used
to generate mosaics. For example, for objects far away from the camera,
traditional mosaic generation methods are used. For incoming images of a
video sequence I1 , $ , I N , a mosaic Mk defined at time k is incrementally com-
posed by combining the mosaic Mk–1 with the current image Ik. Thus, in
parallel to the video sequence, we can define a (partial) mosaic
sequence M1 ,$ , MN . The spatial dimensions of this mosaic sequence change
for each mosaic with time. The art of this composition has been explored in
the last 25 years by using different blending techniques, i.e., the composition
of Mk given Mk–1 and Ik. Traditional application areas were astronomical,
biological, and surveillance data. The need of more refined methods, which
are used in robotics-related applications or immersive multimedia environ-
ments, required the processing of detailed 2-D and/or 3-D visual informa-
tion, e.g., the segmentation of 2-D and/or 3-D objects. In this case, just
knowledge of pure 2-D image information is not enough. We describe next
how mosaics are generated in this case.

2Parallax describes the relative motion in the image plane of the projection of objects in (3-D)
scenes: objects closer to the camera move at higher speed than objects further away.

© 2005 by CRC Press LLC


Exploring the Variety of Random
Documents with Different Content
 CHAPTER VII.
A FAMILY LIKENESS.

There was a sense of mystery hanging over the Castle. Visitors were
expected, and a whole suite of apartments, called the Queen’s
Rooms (of course because Queen Elizabeth had occupied them on
one of her progresses), had been prepared for them, much to
Félicia’s astonishment and somewhat to her indignation. It seemed
to her that she and her engagement had sunk into insignificance in
view of this approaching visit of Lord and Lady Cyril Mortimer and
the latter’s son. Why on earth should such a fuss be made about
them? she wished to know; and she took refuge in an insulted
determination not to show any interest in their coming, although
Maimie’s eyes expressed abundant knowledge, and Usk was
obviously willing to be questioned.
Lord Caerleon and his son drove into Aberkerran to meet the
travellers, and afternoon tea was postponed until their return. This
was a fresh grievance for Félicia, who declared herself as limp as a
rag. In the intervals of eyeing the tea-table thirstily, she spared a
little mild wonder for Lady Caerleon, who was moving nervously
about the hall, altering the position of the furniture and rearranging
the folds of curtains. The brougham drove up, and Lady Caerleon
flew to the door to receive the white-haired lady whom her husband
led up the steps. Félicia’s quick eye noticed at once that the visitor
was wearing the most magnificent sables she had ever beheld, but
her amazement was extreme when she saw Lady Caerleon only
checked in a deep reverence by the newcomer’s seizing her hands
and kissing her on both cheeks.
“Do tell, now!” whispered the astonished observer eagerly to
Maimie. “Aren’t these English people real stiff? or is it because Lady
Caerleon is a foreigner?”
“Nadia, my dear sister!” the visitor was saying reproachfully, as
Maimie shot a glance of scorn at Félicia, “you will astonish your
young friends. Present them, please—I mean, you will introduce
them, won’t you?”
As Lady Caerleon complied with the request, in a curiously
flurried manner, Félicia noticed that the stranger’s face did not look
old, in spite of her white hair, and that she had a very sweet smile.
In her bearing there was something so dignified, notwithstanding
the gentleness of her words and looks, that Félicia felt as if she was
being presented at Court. In Lady Cyril’s presence she suddenly saw
herself as an outsider, and began to feel nervous about her manners
and uncomfortable about her voice.
“But I’ll watch and see how she fixes things,” she told herself,
with returning confidence. “I guess an American woman can just
make herself over if she wants to, and not be beaten by any Eu-
ropian court lady.”
Lord Cyril’s greeting restored her self-complacency. He was “just
ordinary,” she decided at a glance, small and grey-haired, with blue
eyes and unobtrusive manners, but he expressed it as his opinion
that Usk was a very lucky fellow, which showed him to be a person
of discernment. He had brought a secretary with him, it appeared,
and his wife a lady companion, but Lady Cyril’s son, Baron von
Neuburg, had been delayed on the Continent by the breakdown of a
train, and would not arrive till the next day. Félicia was still
oppressed by a sense of mystery. Why did Lady Caerleon seem to
leave the initiative in everything to her sister-in-law, and why was
Lady Cyril constantly on the point of taking the lead in another
person’s house, to the obvious amusement of her husband? The
secretary and the lady companion, who were introduced as M.
Paschics and Mlle. Mirkovics, were stiff and silent, and appeared, no
doubt unintentionally, to disapprove of what was going on; and,
most unpardonable of all, Usk’s attention seemed to be devoted
chiefly to his aunt, and not to his betrothed. It was not until tea was
over, and the visitors had been shown to their rooms, that Félicia
was able to mark her sense of his behaviour by refusing to come
into the garden with him to see if there were any primroses out.
Instead, she went upstairs with Maimie.
 “Well!” she said, when they were in their own room, “I would
like to know what’s come to all of the folks. Lady Cyril’s only a
dowdy old thing after all, though her sables are too sweet for words,
but they all treat her just like a queen.”
“And she is a queen,” said Maimie impatiently, “and her sables
were given her by the Emperor of Scythia. She is Queen Ernestine of
Thracia, who married Count Mortimer two years back, as Mr Hicks
has told you time and time again, and her son, who comes here
incognito to-morrow as Baron von Neuburg, is the King of Thracia. I
meant you to marry him.”
“Land alive!” was all that Félicia could say.
“I wanted to make you an empress, and you threw away your
chance,” went on Maimie bitterly. “And this time I was going to have
you marry a king, and a week before he comes you conclude to get
engaged to Usk!”
“Why, Maimie Logan, have you lost your mind?” cried Félicia,
recovering herself. “You know we would never be let do it.”
“I know you just would, then. I could fix it.”
“Well, it’s too late now, and I don’t care, any way. Usk is a real
good fellow, and his folks are just lovely. I’m on yet.”
“I do wonder how you’ll feel when you come to live way down
in the country all of the year, and be out in the air mornings and
afternoons and evenings. You’ll have a nice round contented face
like the Jones girls by then, rosy-cheeked and blowsy and
dairymaidish.”
“You needn’t be nasty.” Félicia went to the glass and examined
her face with some anxiety, but the pure creamy pallor was not yet
vulgarised by any touch of red. She laughed. “Not much harm done
this far, Maimie. I’ll tell you when I’m tired of Usk, any way.”
“No, you won’t,” said Maimie calmly. “You’re tired of him already.
It’s getting more and more of a trial to have him stick to you the way
he does, but you let him do it just to make me mad. When you’re
tired clear out, you can let me know.”
“And you can round up a few of the kings you keep on hand,
and have me choose,” responded Félicia.
 Now that her perplexity was at an end, Félicia was prepared to
take full advantage of the situation. Queen Ernestine’s obvious
difficulty in masquerading in the English household as a younger
son’s wife afforded her a malicious amusement, and she could not
pardon Usk for refusing to respond to the frequent glances she
flashed at him. At the same time, she watched the royal lady
narrowly, hoping to discover the source of the peculiarly dignified
charm of manner which characterised her. Those who knew could
have told Félicia that it was the outcome of a life of sadness and
self-repression, crowned at last with a tardy happiness chastened by
apprehension. Her own soul, however, was not sufficiently awake to
note more than that the Queen seemed to do and say the right thing
by instinct, and that the grace of her bearing was only equalled by
her consideration for others. Knowing little of her story, Félicia was
captivated by her personality—although the attraction was no bar to
the entertainment she derived from seeing her in a false position—
and she felt quite virtuously indignant when Maimie, by what
seemed an unpardonable piece of gaucherie, brought a shadow into
the beautiful changeful eyes.
It appeared as though a demon of mischief had taken
possession of Maimie that evening. The secretary and Mlle. Mirkovics
had discreetly excused themselves after dinner, leaving the family
party alone, and the four elders were gathered by the fire, talking.
Usk and Félicia occupied a three-cornered settee at a little distance—
it was a piece of furniture Usk liked, because the form of the seat
obliged him to turn round and rest his arm on the back, if he was to
face Félicia in talking to her. Between the two groups, at a little table
close to the standard lamp, sat Maimie, looking through the annual
volume of an illustrated paper some twenty or thirty years old, which
she had disinterred from the library for reasons best known to
herself. Occasionally she interjected remarks into the conversation of
the party by the fire, enjoying the delightful feeling that she was
outraging all etiquette, and yet that no one could rebuke her. It
happened that the talk turned upon the escapade of a young
member of a German princely family, who had recently disappeared
from his home, and was understood to have managed to join the
forces opposed to Great Britain in a little war which was then raging.
“Well, now, that’s queer!” said Maimie. “Here in this old paper
there’s a case of the same sort. Just listen: ‘In connection with the
ill-fated Prince Joseph of Arragon, a view of whose yacht, the
Claudine, we published a fortnight ago, a Vindobona correspondent
sends us a romantic story. It is understood that the official accounts
of the Prince’s voyage to the South Seas were merely a blind, and
that his journey was in reality a flight, in which he was accompanied
by a charming young lady of noble birth, whose musical
performances have been the delight of Vindobona this winter. It is
even rumoured that they were privately married. However this may
be, it is certain that not only Fräulein von Lilienkranz, but her
duenna, Frau Schlesinger, and the latter’s daughter Julie, who was
her constant attendant, disappeared at the exact time that the
Prince’s voyage was announced. The keenest interest was felt here
as to the dénoûment of the romance which has been so tragically
ended by the Australian billows——.’ Do tell, Fay!” Maimie broke off
shrilly. “Julie Schlesinger! mightn’t that be my mother’s name? It
caught my eye right away.”
“See if there’s a portrait,” suggested Félicia. “Maybe it was your
mother in disguise, though I don’t just see how it works out.”
“No, there isn’t any. It says you couldn’t buy one in Vindobona—
that either the Prince or his family had snapped all of them up, both
of him and the lady, and destroyed them. There’s just the ship, two
weeks back.”
Usk sprang up, as though to look at the picture, but when he
was stooping over the book he whispered hastily, “Perhaps you don’t
know that Prince Joseph was my aunt’s cousin?”
“Oh, I’m real sorry. How could I know?” Maimie whispered back,
with a face of guileless innocence touched with anguish. Aloud, she
made some remark about another picture on the same page, and
Usk returned to his former seat. As he did so, he was struck by a
certain alert look, which he knew well, on his uncle’s face. Félicia
had turned her head when Maimie claimed her attention, and the
lamplight fell full upon her profile, which was in a line with that of
Queen Ernestine. Usk saw the two faces from one side, his uncle
from the other, but at the same moment it flashed upon both of
them that the profiles were extraordinarily alike. Lord Cyril cast a
glance of calm scrutiny at his nephew, and resumed the
conversation Maimie had interrupted. It was possible that he
regarded the likeness as merely a coincidence, and Usk hoped this
might be the case.
It might have seemed to strengthen this opinion that Lord Cyril
made no remark to his nephew, but some hours later, when he
looked into his wife’s boudoir, where she was spending a few
minutes in chat with the sallow-faced lady-in-waiting, it was clear
that he had not forgotten what he had seen.
“And what do you think of your niece that is to be, Ernestine?”
he asked.
“She is very beautiful,” said the Queen, somewhat doubtfully,
“but it does not seem to me that poor Usk has found her heart yet.”
“Perhaps she has none to find. What does Mlle. Mirkovics
think?”
“Miss Steinherz has something of the grand air, though there is
a want of repose in her manners,” was the temperate reply. “One
would scarcely expect that in an American, perhaps. But she carries
herself like a queen in her own right.”
“Every American woman is that by right of birth,” said Lord Cyril
lazily. “Does Félicia’s face remind you of any one, Ernestine?”
“Oh, you have noticed it, then? The likeness haunted me all
evening, until I happened to see her refusing to say good night to
Usk. He had offended her in some way, and she would only let him
kiss her hand. Then I saw a likeness to my aunt Claudine, who
married the King of Cantabria. It was not a happy marriage, you
know, Cyril, and her portrait has a look of haughty resignation about
it—a kind of ‘scorn of scorn.’ It is only in profile that there is a
likeness, for none of our family have that little rosebud mouth or
those surprised eyebrows, and Félicia is much prettier than my aunt
ever was.”
“It’s a curious coincidence,” said Cyril, apparently dismissing the
subject as he left the room, but his mind was still busy with it.
“There are more coincidences in the affair than this one, if I am not
mistaken,” he said to himself. “There was that story which came out
at the inquest, that the girl’s father was murdered by mistake for the
Archduke Ferdinand Joachim. Come, this narrows it down. Weldart
and Hohenstaufen! King Paul of Cantabria married a Weldart, and his
mother was a Hohenstaufen. The girl shows no traces of Albret or
Hohenstaufen features—happily for herself—but the Weldart look is
very distinct, only fined and sharpened as European types so often
are after a generation or two in America. Clearly, then, the deceased
‘Mr Steinherz’ was a son of Paul and Claudine. There were three of
them. Ramon is still alive, as I know only too well. Florian died
young—in somewhat discreditable circumstances, but he is
indubitably dead. There remains the lost Joseph, and here, I think,
we must look for the link. That shipwreck presents great
possibilities. It looks a little fishy that Prince Joseph and his fair
musician—oh, and the maid too, of course—should have contrived to
escape when every one else on board was drowned. Was there
collusion somewhere? I don’t see how or why the thing was worked.
But perhaps they had trans-shipped before the end came? That’s
more likely, and makes the whole thing credible. Well, Usk knows
whatever there is to be known, and if I am at all acquainted with
him, will make a clean breast of the whole matter to me at the
earliest opportunity. Caerleon and Nadia know nothing, nor does the
girl herself. She is too self-conscious to be a good actress. But the
hanger-on—does she know? or was that artless discovery another
coincidence? She also has possibilities. She allowed herself to be
silenced by Usk, which she would hardly have done if she had been
the brainless innocent she was impersonating for the moment. I am
inclined to think she does know, and is keeping the thing dark for
some reason of her own. What that may be is not apparent for the
moment, but I think it is distinctly to the advantage of all concerned
that silence should still be kept. To-morrow I shall probably hear
what Usk thinks about it.”
The confidence which Cyril anticipated he received the next
morning, when Usk and he were walking into Aberkerran. Queen
Ernestine had invited her sister-in-law and the two girls to assist at
the unpacking of a box of Eastern embroideries she had brought
with her from her Syrian home, and Usk seized the opportunity of
obtaining a private talk with his uncle, who had some telegrams to
send off, and asking his advice. He told his tale as briefly as possible,
anxious to make an end of a disagreeable task, and there were
several points on which Cyril was obliged to seek further information
by means of questions. When it was all clear in his mind, he walked
on for a short time in silence.
“Is there anything you think I ought to do?” asked Usk at last.
“Do? What can you do? You are absolutely debarred from
enlightening Félicia, as I understand, or even from informing any
one else of the circumstances. Obviously, then, the only thing that
practically concerns you is the marriage at St Mary Windicotes.
That’s your starting-point. I suppose you didn’t think of running
down there to look up the register before you went to America?”
“No; I had no opportunity.”
“Do it when you are next in town, and note especially whether
there is that slip of paper pasted on the inside of the cover, or not. It
is just possible that Steinherz père made up the whole story on the
strength of his likeness to the Archduke, though I allow that would
not account for Félicia’s profile. That marriage once established, you
are safe in England and America, at any rate, for people know better
than to inquire too curiously into the pedigree of heiresses from the
States. Of course it would not hold for a moment in Pannonia. There
you would have to take your stand on the quasi-ceremony which
Steinherz devised to spite his chaplain. If he had lived, it would
almost certainly have been annulled on an appeal to Rome by his
family, but as he was otherwise disposed of, no doubt they thought
the less fuss made the better. As it is, you see, the sole evidence for
it is the word of a dead man, for the other witnesses would know
better than to testify to it unless the Emperor directed them to open
their lips.”
“But would it in any circumstance be valid?”
“It is the kind of marriage that may be either valid or invalid at
the will of the families concerned. If it was to the advantage of the
Albrets to consider it an ordinary morganatic marriage, they would
condone Steinherz’s defiance, and probably induce the Emperor to
confer a title of nobility on Félicia. Or if there was something really
important at stake—for instance, if you took it into your head to
revive your father’s claim to Thracia, and made yourself troublesome
in the Balkans, and they found they must buy you off—they might
even be brought to recognise the marriage as fully valid and regular,
and declare Félicia a Princess of Arragon, just as a bribe to you to
stay at home. It has been done, notably about two hundred years
ago, in a case intimately connected with English history, but
naturally, the consideration would have to be a large one. And on
the other hand, if they see no advantage to them in the matter, they
can deny the Pannonian marriage, and refuse to recognise the
English one, and continue to give out officially that Prince Joseph
was drowned off the Australian coast, however forcibly it may be
proved to them that he wasn’t.”
“I’m glad to hear you say this, for I still feel sometimes as if I
was keeping Félicia out of something that was her due.”
“You needn’t. By the bye, I suppose Hicks will come to England
for your wedding, about this trustee-business? I should like to go
through these proofs which he has in charge. If I am to be co-
trustee, I must know where we stand. But I don’t think you need be
afraid the thing will ever come out. You won’t haunt the Pannonian
Court, I presume?”
“No, indeed!” said Usk fervently. “England for me!”
“If Félicia will allow it. Ah, that reminds me. Is the companion—
the Logan girl—supposed to know all this?”
“Miss Logan? Certainly not. I don’t think anything would have
induced Mr Steinherz to tell her.”
“Still, that is no bar to her having discovered the facts for
herself. I believe she knows, but I don’t quite see what use she
means to make of her knowledge. I could wish that King Michael
was not coming here, though. Is it necessary to warn you not to let
Félicia see too much of him?”
“I wouldn’t spy upon her on any account.”
“Quite unnecessary, if you stick to her as closely as you do now.
No one else could get a word with her. But it’s just possible she may
get a little tired of your constant vigilance.”
“There’s no vigilance in the matter,” said Usk warmly; “it’s simply
that I like to be with her, of course. And if I didn’t think she liked it
as much as I do, I’d—I’d try to keep away from her a little.”
“Don’t imagine that I am blaming you or assailing her. I think I
have mastered the art of letting people be miserable in their own
way. If you prefer to be unhappy with Miss Steinherz rather than
without her, far be it from me to interfere, even though your taste
may surprise me as much as my own wife’s does. I merely advise
you to remember that Michael has a truly royal eye for a pretty face,
and that Félicia may not know the exact value of compliments and
asseverations from a man in his position.”
“What’s bringing him here?” growled Usk.
“Pecuniary difficulties, of course. The last Drakovics Ministry left
the country plunged in debt, and old Mirkovics, who is as honest as
the day, but no financier, made no attempt to retrench. He thought
that anything spent on the army or on public works must be so
much to the good, while the King felt it his duty to look after the
beautifying of the capital and the formation of the public taste, and
between them they’ve called down the thunderbolt which used to
hang over our heads in the old days.”
“You mean they’ve encroached upon the interest due on the
Scythian loan?”
“Encroached? They haven’t paid any interest for three years. Of
course Scythia lay low and accepted their excuses with a pleased
smile, and equally, of course, she is demanding the arrears now.”
“But what can they do?”
“Oh, Scythia is most willing to suggest terms—and conditions
also. One of the conditions is a Scythian marriage for Michael, and
the others would deliver Thracia over, bound hand and foot, to
Scythian influence. Pannonia and Hercynia are not much inclined to
help, having seen all their warnings disregarded, but they will hardly
be able to remain passive in view of such a complete surrender. The
problem of the moment is to secure their moral support without
requiring them to advance the corresponding cash, and King Michael
is coming here in the hope of solving it.”
 “I see. He has got himself into difficulties, and expects you to
get him out of them?”
“Quite so. As soon as he found himself in this fix, he
telegraphed to request me to return and reorganise Thracia. It
showed a touching faith in me; and once, no doubt, I could have
done what he expects of me, but not now. All I can do is to go
through his affairs, and give him the advice of an ordinary business
man—no brilliant strokes at this time of day. I know he thinks I can
raise the necessary money among my Jewish friends by my personal
influence, but they have something else on hand.”
“But do you mean,” cried Usk indignantly, “that King Michael
insisted on your coming back to Europe in spite of the risk, and that
is why you are here?”
“You forget that he’s my wife’s son. It pleases his mother when
he turns to me for help, and I am glad to do what I can for him. But
I had other reasons for coming home. There is one of them.” He
pointed to two men, armed with fishing-rods and baskets, who
passed at the moment, saluting Usk with marked deference.
“But those are the detectives who are down here to look after
King Michael,” said Usk. “The stout one comes from Scotland Yard,
and is sent by our Government, and the other, who belongs to some
private office, is employed by the Thracian Minister.”
“He may belong to any number of private offices, but he is
employed by the Scythian Secret Police, and he is here to spy upon
me.”
“And my father has given him free leave to go anywhere he likes
on our land, and to fish, and so on, and he has taken up his quarters
in the village!” cried Usk, aghast. “But we’ll soon kick him out.”
“I beg you won’t do anything of the kind. His presence here is a
testimony to my former importance, which I find very soothing.
Moreover, he is to be made useful. You saw those cipher telegrams I
sent off just now? They all have to do with a Zionist scheme for
making me Prince of Palestine, and while I am here I shall be
continually having letters and telegrams and even visitors, all
connected with the same thing.”
“But this fellow will be spying about and spoil everything.”
 “He will see and report everything, for that is what he is meant
to do, but he will spoil nothing. The whole thing is a blind, intended
just to keep the spies busy, and the rank and file of the Children of
Zion quiet, for fear of their doing something rash. The real plot is a
very different matter. You may as well know the main idea. The
United Nation Syndicate has been reconstituted, but its object is to
be attained in a new way. So long as Scythia holds Jerusalem, the
Jewish ideal is necessarily incapable of realisation. She cannot be
dispossessed by force; can diplomacy do anything? Besides the Jews
there is another very powerful body whose interest it is to get her
out, and both these communities have fixed upon a certain person
as able to do it if any one can. No; I am not the happy man. My part
is merely to cover up his tracks.”
“But the risk!” cried Usk. “Why should you run into danger for
the sake of this other man?”
“To do old friends a good turn, I suppose. Or perhaps, now that
I can’t play the game myself, it’s the next best thing to see other
people doing it well. Have you guessed who the man is? It’s
Malasorte, the Neustrian Pretender. He is to be assisted to make
himself emperor at last.”
“But I don’t see——How will that——?” Usk was unwontedly
perturbed.
“The Jesuits will arrange things for him at home, the Jews
abroad. Moral support and ready cash, the two indispensables again.
In return for their help, he has promised the Jews to turn Scythia
out of Palestine.”
“And what has he promised the Jesuits?”
“Ah, that I don’t know, and there, it seems to me, is the weak
point. What if the Jesuits also want Jerusalem? If so, he may hand it
over to the Jews first, and then kick them out of it for the benefit of
the other lot. ‘Mala sorte, buona fede’ is the family motto, you know,
but no one ever thinks of quoting it except as ‘Buona sorte, mala
fede.’”
“But why do the Jews and the Jesuits both think so much of
him?”
 “He is a very unusual man, no one can doubt it. His enemies are
fond of asserting that he is not a genuine Malasorte, but his likeness
to Timoleon I. is enough to disprove that calumny. His father was in
the Scythian service, and married a Greek lady, descended, of
course, from the Byzantine Emperors. That gives him a great pull
with the Orthodox, and he is a personal friend of the Emperor of
Scythia, who is very much under his influence. He believes he can
induce him to withdraw his troops from Palestine, and Goldberg and
the rest believe it too. So he is to have his chance.”
“It’s awfully queer how things are mixed up,” said Usk. “Do you
know this chap wanted to marry Félicia?”
“This same man—Timoleon Lucanor? Yes, I remember, he was
military attaché at Washington. And why didn’t he?”
“Her father didn’t think his chances good enough, I fancy.”
“No; I think the blood of the Albrets rose against the parvenu,
even then. But this is interesting, Usk. Could Malasorte have had any
idea of Mr Steinherz’s secret? No, of course not. He would not have
allowed himself to be choked off in that case. It was the money he
was after, and he missed what would have been better to him than
millions. Did Miss Logan approve of his suit, do you know?”
“Yes; Hicks said she pushed it all she knew.”
“And yet she didn’t tell him? It’s pretty clear, then, that
whatever she knows she has learnt since. My dear Usk, you may
thank your stars that I am a wretched crack-brained failure, if you
want to keep your Félicia. There are the makings of a most
tremendous plot lying about—infinite possibilities—and I can’t see
how to put them together.”
 CHAPTER VIII.
LOVE IN IDLENESS.

The arrival of Baron von Neuburg, otherwise King Michael of Thracia,

did not add to the gaiety of the circle at the Castle. Scarcely more
than a boy in years, his face was so curiously old that not only Usk
but his father looked young beside him, and his manner was weary
almost to the point of exhaustion.
“I don’t take much stock in kings generally,” Félicia remarked to
Usk after the new-comer had been presented to her; “but for the
sake of the rest I hope this one’s a bad sample.”
“He’s not a particularly good specimen, certainly. Aren’t his ways
awfully riling?”
“Yes,” agreed Félicia; “that’s where the difference comes in
between him and your uncle. Lord Cyril is just elegant. He has
dipped into everything, and got pretty tired, but he don’t advertise
the fact. The Baron has done it too, and parades his weariness, and
that’s rude.”
“I’m glad you don’t care for him,” said Usk honestly. “He’s not—
not the sort of fellow I should like to see you take to.”
“Don’t you know that by saying that you’re just daring me to be
as sweet to him as I know how?”
“Oh, I know people say that sort of thing about women, but no
nice woman would go and make up to a man of bad character
simply because she was warned against him.”
“Then I’m a nice woman? Well, I guess you’ll expect me do
something for you in return for that acknowledgment, and I’d admire
to teach the Baron a lesson.”
“Not by way of breaking his heart, please.”
“Not while you’re around, any way. But I’d like to have him
know what folks would think of him if he didn’t just happen to be a
king.”
 It is possible that Félicia set to work with all the more gusto that
she was conscious of a personal injury at her introduction to the
King. To her mind his look had expressed no recognition of the fact
that he was being presented to a very beautiful woman. Maimie also
had noticed this insensibility, and she commented upon it to Félicia,
with a certain lack of tact.
“You see, Fay, he can just tolerate having you around as the
future Lady Usk, but it’s quite beyond him to show any interest in an
untitled American girl. She’s way down under his feet in the mud
somewhere.”
“Well, I guess it’ll maybe interest him to know what the
American girl thinks about him.”
“Don’t see how you’re going to have him feel it, any way. Say,
Fay, Usk is real devoted, isn’t he? I’m glad you’ve taken him instead
of the King now I’ve seen him, don’t you think! He’s solid good right
through. He won’t ever have you find out anything new and
unpleasant about him. Every day of his life he’ll come to meet you
mornings with a flower, just as his father does to Lady Caerleon, and
he’ll like nothing better evenings than sit alone with you and read
the ‘Times’ out loud.”
“Guess I’ll fix things differently to that.”
“You’d better not. His way will save you a pretty good deal of
trouble. He’ll like you to be the same all the time, just as he is. And
so’s you’re just decently civil to him, he’ll never be ugly. It’s only with
a man like the Baron you need to be smart, for you lose him if he’s
away out of your sight a moment.”
“Well, I incline to think the Baron will need to be smart this next
week.”
In all probability the Baron was of the same opinion before the
week was over. He had come prepared to take the lead in the
general conversation, even if he did not monopolise it, and nothing
was further from his thoughts than that this little nobody of an
American, whom the Caerleons had managed to pick up for Usk on
account of her money, should take it upon herself to dispute his
right. Young as he was, King Michael had already ruled so long that
it seemed to him only natural to be the autocrat of any table at
which he sat, and it was whispered that his meetings with other
monarchs were few and far between, and also extremely short, by
reason of this genial habit of mind. It was a tremendous shock to
find himself called upon for explanations, laughed at for his choice of
words, even contradicted, and all by a radiant being who flashed
provoking glances at him from magnificent eyes, and having
annihilated him, turned with irritating nonchalance to engage Usk in
a low-toned conversation punctuated with soft looks.
To Usk himself these favours, thus publicly conferred, were the
reverse of delightful. His hospitable soul was wounded by the
treatment meted out to the guest, and he could not help feeling that
it was bad form in Félicia to emphasise his own happier position at
every opportunity. When he ventured to remonstrate, however, his
only reward was a severe snubbing in private, for Félicia was not to
be turned from her prey. Lord Cyril, to whom Lady Caerleon
appealed in distress, laughed at the whole thing, and declared that
Félicia’s scorn was the best possible tonic for his stepson. To have
met her would be a liberal education for him, provided she continued
the treatment to the end of his visit, and did not soften towards him
for a moment. But there was no need for her to do this. It was
enough for King Michael to notice the difference in her manner when
she turned to Usk. When a rarely beautiful woman treats one man to
nothing but gibes, and lavishes tenderness upon another who looks
rather uncomfortable under the process, it is not in human nature
not to wish to be in the other man’s place for once.
Maimie had been watching eagerly for the King to reach this
point, but here she found herself at a standstill. All this time she had
been cultivating the acquaintance of the aide-de-camp, Captain
Andreivics, who had accompanied King Michael, and was introduced
as his “friend.” She had picked his brains to such good purpose that
she knew as much as he did of the King’s circumstances, and had
gone far beyond him in the deductions she drew. He could not tell
her exactly how the long hours were spent when the King and his
stepfather were closeted every morning with vast piles of papers,
but she knew. She could picture Cyril exposing pitilessly the
extravagance, laxity, and corruption which had spread through every
department of State since he had left Thracia, and indicating reforms
and economies which would put matters straight if the present crisis
could be tided over, but performing no miracle to provide the money
urgently needed at the moment. It was from the aide-de-camp,
however, that she heard how one morning the King dashed away the
papers with the pettish remark, “Really, Count, your brain seems as
strong as ever for all these absurd trivialities. Why is it that you
refuse to return to Thracia and get me out of my difficulty?” and
how the incisive answer flashed forth, “If my brain was as strong as
ever, sir, I should not be busy with these trivialities. I should be
holding the balance of power in Europe.” She knew that both Cyril
and the King had reluctantly come to the conclusion that, failing the
much-needed miracle, there was only one thing to be done. The
King must journey homewards by way of the Riviera, where the
Scythian Princess who had been proposed for his acceptance was
sojourning. There was a bare hope that his apparent intention of
falling in with the arrangement suggested by Scythia might alarm his
Pannonian and Hercynian relations into some attempt to prevent the
threatened surrender.
Maimie ground her teeth, metaphorically speaking, over this
deadlock. There were Félicia’s millions, far more than sufficient to fill
the yawning gulf, for the sum which could bring a Balkan State to
bankruptcy was trivial in American eyes, and no one seemed to have
thought of making use of them. The King, however much he might
admire the girl of whose beauty he had at last become conscious,
had not the slightest thought of marrying her, and Cyril would take
no step to utilise those convenient millions, even if they ever
occurred to him, because Félicia happened to be engaged to Usk.
Maimie felt that she had no patience with Usk, who to the injury of
existing added insult by hanging about Félicia so perpetually that the
King had no opportunity of getting up a flirtation with her even if he
desired it. If Usk would only run up to town for a day or two! but he
remained in Félicia’s near neighbourhood as persistently as if he had
known of Maimie’s designs and meant to thwart them. Captain
Andreivics it was who returned to London two days before the time
fixed for the King’s departure, and with no better excuse than the
stereotyped one of “urgent business,” so that Maimie lost her
cavalier.
At last, just when Maimie was gloomily revolving in her mind
various desperate expedients for removing Usk from Félicia’s side,
the motive force required was suddenly imparted from without. It
came in the form of a letter from Mr Forfar, the Prime Minister, to
Lord Caerleon, asking whether Usk was still thinking of entering
Parliament. If so, he could do a great service to the party (to which
the Marquis had always lent a loyal, if discriminating support, since
his own entry on public life), by allowing himself to be adopted as
the future candidate for a great Northern constituency. The sitting
member was old and feeble, but had stoutly refused to tolerate the
mention of a successor until recently, when a severe illness had
given him a fright. He was now willing to allow a suitable “under-
study” to be introduced to the party managers and make himself
known in the constituency; and as the supporters of the Temperance
cause were well organised and powerful, what more suitable
candidate could be found than the son of the life-long Temperance
champion? Neither Usk nor his father hesitated a moment in
accepting the offer. The great banquet, at which Sir James Morrell
had reluctantly undertaken to present his successor to the
association which had so often shared with him the sweets of
victory, was to take place two nights after the arrival of Mr Forfar’s
letter, and there was no time to be lost. Usk dashed upstairs in high
excitement to pack his bag, while his father went round to the
stables himself to order the dogcart, and Lady Caerleon interviewed
the cook on the subject of sandwiches. It never entered Usk’s mind
that any one could dream he would let slip this long-desired
opportunity, and he sent an eager message by Maimie begging
Félicia, who was breakfasting in her own room, to drive to the
station with him. When he came downstairs with his bag, however,
he found her still in her “wrapper,” as she called the frilled and
beribboned garment which Lady Caerleon always felt ought not to
make its appearance outside a bedroom. Maimie had insisted on her
getting up when she brought Usk’s message, and she was obliged to
make a very hasty toilet, much to her disgust. She did not appear to
suitable advantage, she felt, unless Pringle’s skilful fingers had
proper time for their work.
“Why, Fay, aren’t you coming, then?” cried Usk, when he saw
her.
“I guess not,” responded Félicia laconically.
“Well, you’ll wish me joy, won’t you? This is the beginning, you
know.”
“Of what? The beginning of the end?”
“What do you mean?” Usk was too happy and too much excited
to make any attempt to understand. “Why, it’s my chance at last—
what we’ve talked of so often.”
“You have,” corrected Félicia. Then a transient gleam of
brightness showed itself. “Canvassing, do you mean? Then I suppose
we are all to trail you along?”
“Oh no, that would be a little previous,” laughed Usk. “I’m only
going to be introduced to the party. There’s no election on at
present.”
“I want to know! You’re leaving me this way just for an ordinary
ward-meeting? and you don’t so much as ask me whether I choose
to have you go?”
“Why in the world should I think you’d mind? I thought you
would be delighted. Why, Fay!” in utter amazement, for Félicia was
weeping delicately into a lace handkerchief.
“You said you loved me, and now you’re having the Baron—
everybody—see how little you care for me. You just haven’t the very
slightest consideration for my feelings!”
Dismayed and astounded, but still utterly puzzled as to the
nature of his offence, Usk knelt down hastily by her chair, and
alternately entreated her forgiveness and adjured her to tell him
what he had done. With him it was a matter of course that the
women of a household should send forth their men to the chances
of war or politics with a brave face and words of cheer, and he could
not conceive Félicia’s feeling hurt at not being consulted. He did not
know that Maimie’s triumphant “Now you see just how much he
thinks about you!” was rankling in her mind, and that she had set
her heart on proving her power over him. When she consented at
last to remove the handkerchief from her eyes, it was merely to
intimate that he might consider himself forgiven if he did not go. Usk
sprang to his feet.
“And lose this chance—give Forfar a slap in the face?” he cried
in astonishment. “Why, Fay, you must be mad! You’re joking, aren’t
you? You couldn’t possibly mean it.”
“I guess I mean exactly what I say. You’re telling me all the time
how much you love me, but I can’t just seem to realise it.”
“You mean that you ask me in cold blood to give up this chance,
disappoint my people, offend Forfar, just because—why, it’s for no
reason at all!”
“I don’t ask you anything,” said Félicia, rising regally, and
throwing him a glance over her shoulder. “I just tell you to do it.”
“Well, then, I won’t,” returned Usk, with equal candour; but as
she swept towards the door he intercepted her, breaking into a
laugh, “Why, Fay, for the moment I thought you meant it. What a
gorgeous scene we have been making over nothing at all! Now you
don’t leave this room until you say you’re sorry, and signify the same
in the usual manner. There! I’m getting quite a public speaker
already.”
“It’s a pity if your future audiences don’t appreciate you better
than the present one,” said Félicia coldly. “Kindly allow me pass. If
you don’t choose to consult my wishes—well, you’re not the only
man in the world, any way.”
“Now it would serve you right if I kept you here until you gave
in,” said Usk, “but I hear my father tramping up and down in furious
anxiety about the train. I’ll settle this little matter with you when I
come back, but any token of penitence in the shape of a letter will
receive due consideration. And—just that you mayn’t make yourself
miserable thinking I’m angry with you—there! and there! and as
many more assurances of pardon as you like.”
He ran along the corridor, still laughing, and Félicia returned
angrily to her own room. It was the laughter that annoyed her. Even
if Usk had refused to yield, he ought to have taken her objection
seriously. He was so sure of her that he thought he could afford to
laugh at her, was he? Very well; it might be advisable to show him
that he need not be quite so sure.
Maimie, in the meantime, had been taking advantage of the
change in the situation. After giving Usk’s message to Félicia, she
wandered down to the second of the terraces before the Castle,
where King Michael was wont to smoke his after-breakfast cigar. It
was his custom to breakfast in his own rooms, and Maimie felt
comfortably certain that Usk would not have thought of hunting for
him in the garden in order to bid him farewell. When she caught
sight of him, he was walking up and down somewhat listlessly, as
though he missed the companionship of Captain Andreivics, but she
saw a change pass over his face when he heard the tap of her heels
on the terrace above. As she came down the steps, she could see
him before he saw her, and it gave her a keen delight to see his look
of disappointment when he met her at the foot.
“Good morning, Baron!” she remarked cheerfully. “Say, we shall
be left awfully lonely to-night, shan’t we—what with the Captain
leaving yesterday, and Lord Usk this morning, and you this evening,
I suppose?”
“I leave to-morrow,” said the King, looking at her with cold
surprise.
“No, is that so? I guess Lord Usk don’t know it, rushing off the
way he has. He would never leave Félicia unguarded a whole day.”
“Is Miss Steinherz supposed to be in danger from me?”
Maimie laughed mischievously. “You know your own reputation
best, Baron. I don’t see but Lord Usk thought it wasn’t enough to
warn Félicia against you, the way he has mounted guard over her.”
“Oh, our friend Usk felt it necessary to warn his bride against
me, did he? I think that was not playing the game, as they say
here.”
“Well, I guess I oughtn’t to have given him away. He knows his
way about, I suppose, and—yes, it was just as well he did it.”
“Why? Has it produced the opposite effect, as usual, and
induced the young lady to honour me with her friendly interest?”
Maimie gave him a glance of compassion. “What good would
there be in that?” she asked curtly. “No; it just showed Félicia what
he expected of her, so’s she concluded to satisfy him at any cost to
herself.”
A light seemed to break upon the King. “What! you mean that
all this raillery, all the contempt she has poured upon me for a whole
week, was nothing but an effort to please her bridegroom?”
“Don’t you try and have me say that Félicia’s an angel,” Maimie
admonished him. “I won’t tell you her secrets, any way. And I don’t
see but you’ll have to stay till to-morrow as you’ve fixed it so; but I
wish you were leaving sooner.”
“Miss Logan’s interest in my movements does me too much
honour. Perhaps it will gratify her to know that it is possible I may
not even be leaving to-morrow.”
“Ah, I thought that ‘urgent business’ of the Captain’s covered
more folks’ affairs than his own. You mean he’ll be coming down
with more documents for you to study with Lord Cyril?”
“I had not meant that——” the King was beginning, but as he
caught the merest hint of scorn in Maimie’s eye, his face assumed an
expression of deep importance. “It is extremely probable. What is
the good of an aide-de-camp but to make himself useful?”
“An aide-de-camp—do tell!” cried Maimie. “And you’ve called
him your friend all the time! Why, you must be a general, then! Say,
General, they promote people pretty young in your country, don’t
they?”
“Mademoiselle,” said the King severely, “with a young lady who
has contrived to discover so much of my private affairs, it is surely
unnecessary to keep up this wearisome farce?”
“M. le Baron,” said Maimie, making him a curtsey, “the farce was
of your own providing. If you choose to throw up your part it can’t
hurt me, any way.”
“You imply that there is another act, if I care to play it?”
“I don’t imply anything. I’m not taking any risks, if you are.”
“So be it. I take the risk. Andreivics shall arrive from London
with important documents. Accept my compliments, mademoiselle. I
sent him away because I wished to feel that a few of my secrets still
remained in my own possession. You have secured his return by
means of a diplomacy which my good stepfather in his best days
might have envied.”
Maimie looked him over with a slow gaze of infinite scorn. “I
guess,” she said calmly, “that you’re sort of acclimatised to being
despised? You seem to lay yourself out for it so naturally. Usk was
pretty wise in warning Félicia against you, and she was doing the
best for herself when she chilled you off. What she can see in you
——”
“Permit me to observe that you are revealing an interesting
secret, mademoiselle,” said the King malignantly. “After what you
have said it would be ungallant in me not to remain here. I telegraph
to Andreivics this morning.”
“Now we begin to be moving!” said Maimie to herself, as she left
him in speechless contempt. “He don’t even see that I’m having him
go the very way I want him. And now for Fay.”
She found that Félicia, having worked herself up into a high
state of resentment against Usk, had determined to punish him by
entering upon a flirtation with the King. Maimie shook her head
when Félicia declared her intentions.
“I wouldn’t, Fay,” she said. “You’ve done elegantly this far, the
way you’ve frozen him off. You’ll only get all of your affairs into a
snarl. And what’s more, I don’t believe he’s to be had. He knows just
how to take care of himself, and I can’t seem to see you making any
sort of impression on him.”
This was all that was needed to put Félicia on her mettle.
“Maimie Logan,” she said decisively, “did you ever know any
man that could take care of himself when I was around? I guess Usk
will be sorry that he went off this way.”
“Why, what do you mean doing?” Maimie’s tone was full of
alarm.
“Oh, just make things uncomfortable a little—nothing more.”
“But if Usk will just stay away four or five days, there’ll be a
good deal more,” was Maimie’s mental comment.

“Michael,” said Queen Ernestine to her son four days later, when
she had succeeded after many vain attempts in finding him alone,
“don’t you think you are paying rather too much attention to Miss
Steinherz?”
She spoke timidly, anticipating the black frown which gathered
at once upon King Michael’s brow, as he bestowed a mental curse
upon Félicia’s methods. It was not in her nature to be content with a
secret adoration. The King might waylay her in the garden if he
chose, or look for her in her favourite nook in the picture-gallery,
and enjoy her society until Maimie, posted judiciously near at hand,
felt it her duty to interrupt them, but he must not attempt to hide his
chains in public. Hence King Michael’s relations had the pleasure of
seeing him dancing attendance upon the whims of a languid beauty,
who had vouchsafed to lay aside much of her sharpness of tongue,
but still betrayed no delight in his attentions. Even Usk could not
have desired more absolute unconsciousness of her conquest than
Félicia exhibited.
“Have you ever known me forget my position?” King Michael
asked at last, when his mother’s face had grown more and more
anxious.
“Never,” she answered, recalling many memories at once
humorous and pathetic.
“Then rest assured that I never shall.”
“But, dear,” urged the Queen, “in that case would it not be as
well to return to London? It looks—mind, I only say it looks—as if
you were taking advantage of Usk’s prolonged absence to rob him of
his bride’s affections, and you would not wish to do that.”
“Count Mortimer has already given me a pretty strong hint to
go, but I will not leave my work undone.”
“But he told me to-day that it was all finished.”
“On the contrary, it is only half-done,” was the reply, with an
enigmatical smile. “Who would have thought I should become so
deeply interested in a matter entered upon so lightly?”
“So lightly—when the future of your kingdom and your own
happiness may depend upon the arrangements you are able to
make?” cried the Queen, in surprise. “You don’t take things seriously
enough, Michael.”
 “Possibly not, but sometimes things take me seriously, quite
against my will.”
********
“I would just love to hustle some folks a little!” Maimie was
reflecting, much about the same time. “Usk will arrive home in a day
or two, and I’d like to have things fixed. I never thought the Baron
would have so much grit in him, but the way he fights off a definite
declaration is real fine. But Fay is even with him there. What with his
precious kingdom, and her engagement, they can’t seem to get on
at all. And until they do, I can’t step in as fairy godmother to put
things straight.”
While the thought was still in her mind, the door opened
violently, and Félicia ran in. Flinging herself upon the sofa, she
began to cry, not weeping in the artistic way which had damped
Usk’s departure, but shedding genuine tears of disappointment and
mortification.
“I’ll never forgive you!” she sobbed out to Maimie. “This is the
second time you’ve put me in just the most horrid sort of a position.
You had me encourage the Malasorte man until he cooled off of his
own accord, and now the King has told me in so many words that he
can’t ask me to marry him—after having me say I cared for him. I
wouldn’t mind so much but for that.”
“Well, you can marry Usk yet,” was the unsympathetic reply.
“After expecting to be a queen!” fresh sobs followed “He spoke
so’s I really concluded it was all safe. And you sit there and say
nothing. I hate you!”
“Why?” asked Maimie calmly.
“How can I help it when you’re so mean and ugly? You told me
you could fix things, and I thought you had fixed them right away.
And then you have all of this happen!”
“Now look here, Fay,” Maimie grasped her shoulder. “If I operate
my scheme right now, will you promise to give up Usk and marry the
King? I won’t go a step without knowing that.”
“I don’t feel like giving him up until you have got things fixed. If
they should chance to go wrong, I would just find myself left.”
 “It may be some time before they go right, I grant that, but you
must take some risks. Well, if I let you stay engaged to Usk for the
present, will you break off with him when I give the word?”
“Ye-es, but you’ll have to be quite sure about it.”
“I’ll see to that. Where did you leave the King?”
“On the second terrace. He was real sorry to have to say what
he did, but I was so mad I wouldn’t stop,” said Félicia, with a curious
kind of self-satisfaction.
“It must be real nice to be able to love folks according to what
they can give you!” soliloquised Maimie, as she went in search of the
King. “She would accept Usk when there was no other man in view,
but now the Baron holds the winning cards—unless there should be
any fascinating emperors around before we get things fixed. But
here’s the lucky man!”
King Michael was walking from end to end of the terrace,
smoking moodily. His hands were thrust deep into his pockets, and
the blackest possible frown was on his brow. He was hard hit,
Maimie saw, and scarcely likely to welcome the appearance of the
person to whom he might consider that he owed his present
unhappy frame of mind, but she met him boldly.
“Say, Baron,” she said, placing herself in his path, “is it true
you’ve told my Félicia you love her, but can’t marry her?”
“It is, mademoiselle. If you have chosen this terrace as a
promenade, I will go elsewhere.”
“Well, if I was a king, I guess I’d marry the woman I loved.”
“Pardon me, I am not only a king. I am a son of the house of
Schwarzwald-Molzau, and we do not mingle the blood of
Charlemagne with that of—manufacturers.”
“And so you sigh as a lover and obey as a son? But you wanted
to marry Lord Usk’s sister once, I know, for Mr Hicks told me.”
The King’s eyes flashed gloomily. “I do not understand this
catechism,” he said angrily. “It is intolerable! The Mortimer blood is
equal to that of any semi-royal house in Europe, and there were
special reasons why a marriage with Lady Philippa would have been
very pleasing to my subjects.”
 “Well, I guess a marriage with a Princess of Arragon would
about satisfy them, any way.”
“Not if she were a Catholic.”
“Is that so? Well, the girl you have just told that you love her
and can’t marry her is a Princess of Arragon and a Protestant.”
“Impossible. I know the Prince of Arragon’s three daughters
well, and Don Florian’s only daughter is married to another cousin of
mine.”
“You are forgetting. There was a third brother.”
“Don José? But he was not married. Oh, I remember there was
some talk of a morganatic marriage. But that is not to the purpose.”
“Excuse me, but I guess it’s very much to the purpose.”
“Allow me to say that you do not know how these things are
regarded—how I regard them, necessarily.”
“But suppose the marriage was recognised? Wait, I’ll tell you
about it,” and she ran through the circumstances hastily. “If the
marriage was good enough for the Arragon family to accept Félicia
as one of themselves, I guess it would be good enough for you?”
“Undoubtedly; but I do not see the faintest likelihood that the
House of Albret would recognise the marriage.”
“That’s where your help would be wanted. I’d like you to lay the
whole thing before Lord Cyril to-day, and have him operate it.”
But the King started back, aghast. “My dear Miss Logan, do you
not see that it would be fatal—suicidal—for me to appear in the
matter? Count Mortimer has his nephew’s interests to consider.”
“That is so; we mustn’t go ahead too fast,” said Maimie. “Then
you incline to think Félicia and I may fight all of the battle for
ourselves?”
“Quite so. There can be no objection to your consulting Count
Mortimer, purely on her behalf, you know; and if by any chance your
efforts should be crowned with success, why—you have given me
hope—I am happy again!”
“Oh no, you aren’t—not yet. We shall want your help any way. I
know well enough there’s no hope of having Félicia’s family
recognise her unless some one puts pressure on them. Are you
ready to intimate at the critical moment that you are real keen on
marrying her, if they can fix things right?”
“That would be highly injudicious. I think I had better not
appear——”
“As you please. You throw up your cards, then, and Félicia will
just marry Usk.”
“You place me in a most difficult position, but rather than lose
Félicia——”
“I thought so. Then you would better go way back to the Riviera
right now, and make a little gentle love to the Grand-Duchess Sonya,
just to keep your folks sort of interested, you know. I’ll let you hear
when you’re wanted.”
 CHAPTER IX.
A CHANGE OF VENUE.

“Why, Uncle Cyril! how awfully good of you to come and meet me!”
Yet Usk’s eyes strayed to the dogcart waiting in the road just beyond
the station fence, and the stolid groom in charge of it. “Is—is any
one else here?”
“No; Félicia is not here. What do you think of putting your bag
into the cart, and walking up?”
“All right. I shall be glad to stretch my legs. Félicia isn’t ill, is
she?”
“I saw no signs of it when I started. But why should you expect
her to meet you? I understood you and she had quarrelled?”
“Quarrelled? Why, it was nothing—the most utter nonsense! She
never wrote me a word for four whole days, though. But I wrote to
her every day, and at last, on the fifth evening, I had a letter from
her—an awfully jolly letter, but making the most tremendous fuss
about the way she had behaved, calling herself names, and all sorts
of things. It seemed so uncalled-for that I really thought she must
be going to be ill, for she’s not a bit morbid generally, is she?”
“Few people less so, I should think. The letter reached you five
days after you left here, you say. It was written the day before, of
course?”
“The evening before. I know she said she was writing when the
house was quiet. But I’m awfully glad she’s all right. She’s so
unexpected, isn’t she? You never can tell what she’ll do next.”
“I used to notice the same thing about my wife in the early days
of our acquaintance. There is a peculiar charm about that
unexpectedness when it is introduced into politics. It quite prevents
any feeling of flatness.”
“Now one would have imagined”—Usk was still pursuing his own
train of thought—“that she would have come to meet me after that
letter.”
 “The unexpected again, you see.”
“But how did you know anything about our——? Well, it wasn’t
a quarrel——”
“The suspension of friendly relations? I inferred it from what I
saw after you were gone.”
“And you spoke to Félicia? Very kind of you, I’m sure, but—well,
you know——”
“You prefer to conduct your own love affairs? Quite so. Make
your mind easy; I did not speak to Félicia. But if I remember rightly,
I did send you in your mother’s letter a strong hint not to stay away
more than the two days you intended at first.”
“Yes, I know, but you said ‘unless it will damage your chances,’
and it would have done, horribly. You see, it was such a piece of
good luck old Morrell’s taking to me so tremendously, when he had
hated the very mention of a successor before, that I couldn’t go and
hurt his feelings. He would drag me round the constituency, and
hunt up all the local organisers to introduce me to them, and we
really covered an immense amount of ground. The party agent said I
couldn’t have made a better start.”
“Don’t think I want to see you less keen. It isn’t that. Did Félicia
tell you any news in her letter—anything that had happened?”
“No; there was nothing of that sort. But really, Uncle Cyril, I
don’t think she was angry with me for staying away. I wrote her
awfully long letters—and sent her things, too. She couldn’t think I
had forgotten her.”
“I never thought she did. But did she express regret for
anything in particular, or merely for her general treatment of you? I
have an object in asking,” as Usk looked at him in surprise. “Don’t
think it’s mere curiosity.”
“She didn’t mention anything definite—except just to say that if
things went wrong between us, it was Maimie Logan’s fault, not
hers, which I could have told her myself. Oh, by the bye, that’s
another queer thing. I had an hour or two to spare in town, so I ran
down to Bradcross and looked up Mr and Mrs Steinherz’s marriage at
St Mary Windicotes. It was there all right, but the queer thing is that