"This three-volume set is a remarkable collection of articles on computer
vision and image processing. Although written by a diverse collection of
experts, the chapters have been superbly edited to achieve a coherent
presentation, with consistent notation and perspective throughout.
Spanning most areas in the field of computer vision, its applications
and its foundations, most of the individual chapters are self-contained.
The editors have also added well-written chapters to provide a solid
base of introductory material.
Computer vision is an interdisciplinary field of research with many
industrial and scientific domains of application. Its foundations are
drawn from physics, signal processing, applied mathematics, computer
science, electrical engineering, and neuroscience. Applications include
mobile robotics, medical image analysis, human-computer interaction,
and scientific areas for which computer vision image analysis techniques
are becoming a key source of scientific data. Indeed, one of the great
strengths of this three-volume set is its focus on both the theoretical
foundations of the field and the applications of the techniques.
The first volume introduces the principles of imaging, which editors
of this book are in a unique position to produce. These chapters
include detailed introductions to optics, radiometry, reflectance,
illumination sources, and sensor technologies. The handbook also
extends this introduction to several other imaging modalities, including
3d range imaging, infrared imaging, MRI, fluoroscopic imaging, and electron
microscopy. These chapters provide a thorough, accessible treatment of
imaging that is unparalleled in any book on computer vision or image
processing. Volume one of the handbook constitutes a definitive and
coherent treatment of radiometry, imaging and sensors.
Volume two introduces the mathematical foundations of image processing
and computer vision. These chapters cover many core areas of computer
vision research (e.g., multiscale image representation, motion, texture,
stereo, shape-from-shading, etc) as well as more recent promising
developments (e.g., nonlinear diffusion, robust variational techniques,
and Bayesian methods). While this volume may not be appropriate for a
course textbook on computer vision, it should provide students with an
outstanding resource.
Volume three is a unique contribution to the computer vision literature;
never before have editors been so dedicated to including thorough, careful
treatments of such a broad range of industrial and scientific applications
of computer vision. With over 40 chapters on applications, there are
articles on photogrammetry, oceanography, botany, surveillance, people
tracking, visually-guided control, 3d object recognition, and many others.
While, such applications are often ignored by textbooks and edited volumes
devoted to computer vision, this unique volume provides a vivid snapshot
of the depth and diversity of the state of the art in computer vision
applications today.
This handbook on computer vision and its applications is a unique,
comprehensive reference that all libraries and computer vision laboratories
will want to have on their shelves. It will also serve as a rich resource
for instructors teaching courses on computer vision and image processing;
the enclosed CD contains pdf versions of all chapters, which will greatly
enhance the value of the handbook.
-David Fleet, Xerox Palo Alto Research Ctr.
These three volumes collect an enormous amount of information about
technical aspect of computer vision, and are written by some of the best
European scientists in the field. Every chapter explores a particular
device, technique, or application in much greater detail, and strives for
greater accuracy of results, than standard computer vision books do.
Applications and examples describe real systems, actually deployed in
industrial and scientific settings.
Imaging systems are studied well beyond the pinhole camera model. A whole
volume is devoted to them, ranging from a thorough and rigorous discussion
of the basic principles, both for two- and three-dimensional acquisition,
through an authoritative analysis of sensors, and to up-to-date accounts of
imagers of all types: cameras, ultrasound and acoustic imagers, range
finders, microscopes, MRI devices, and more.
Volume two covers "classical" computer vision techniques. The vast range of
topics, from signal processing to knowledge-based interpretation of images,
is covered with unprecedented thoroughness, and a fantastic effort has been
made to treat all topics in a uniform, simple, and well designed notation,
and to place them into a coherent conceptual framework. Individual chapters,
or collections of chapters, from volume two can be used as an excellent
basis for college courses on nearly every aspect of computer vision. The
accompanying CD-ROM, with its wealth of both images and software, can add
substantial value to such courses.
Volume three, being a collection of case studies, is more varied in its
level of coverage, with some of the chapters describing preliminary results,
and others, like Dickmann's chapter on dynamic vision, surveying decades of
concentrated work on a particular topic. In all cases, however, attention
for detail and strive for accuracy set this collection apart.
Together, these three impressive volumes testify to the maturing role of
computer vision as a discipline with both conceptual depth and great
industrial, scientific, and technical relevance."
-Carlo Tomasi, Robotics Laboratory, Stanford University
