Издательство Oxford University Press, 2000, -322 pp.Image processing, image analysis, and pattern recognition are techniques now widely used in bioscience and medicine. There is a plethora of packages and systems which one can buy, or download free, ranging from menu-driven systems to libraries for C-programming. There is also a large number of text- textbooks, with one or more of the above terms in their titles, which explain the computational basis of these techniques. Many of these textbooks are excellent in their mathematical and computational descriptions, but take as their audience engineers and computer scientists for whom the methods themselves present the scientific interest. This volume, in common with others in the Practical Approach series, takes a different perspective. We take our audience to be scientists whose interests are in other fields, and for whom these methods provide useful analytical tools. As is the case with all tools it is best for the user to have an idea how they work, so that their behaviour can be understood, without necessarily becoming immersed in the details. As this volume forms part of a series aimed at biologists, we have retained that emphasis. However, the methods we describe are equally applicable in many other fields—the earth sciences for example—and we hope it will be of interest to scientists in these areas also. Try to imagine your favourite images in the place of the chromosomes or fungal mycelia. As in any field of engineering, there are carefully worked out computational and mathematical principles in image processing and analysis, from which practitioners have developed a collection of rules of thumb and established common procedures. It is the latter aspect, rather than the former, that we hope to emphasize here. Some of these procedures take the form of well-known algorithms, and these, together with rules of thumb and standard pro- processes, are presented as Protocols. Some standard pieces of knowledge are more readily expressed in mathematical formulae. This is a mathematical topic, and it would be impossible to describe it without the use of mathematical language. We understand, however, that mathematics does not hold a deep fascination for at least some of our target audience, and we have tried to keep the mathematical descriptions at a level that would be acceptable to a final-year high-school or first-year university student. Many of the algorithms and formulae are already implemented in software packages and libraries. We hope that the descriptions here will give the user some understanding of the reasoning behind the functions given and the limitations of their operation. The algorithms should be presented in sufficient detail that a competent programmer, without much image analysis experience, should be able to implement them if required. There is a wide range of material in the image processing and analysis litera- literature. We have deliberately set out to avoid producing another all-encompassing book. The material we have chosen has two aims. First, it is intended to cover the basic methods of which a user of the technology should be aware. Second, we present a selection of more advanced techniques that we hope will inspire the application of image analysis to a wider range of complex scientific prob- problems. Each chapter describes in some detail a specific technique or collection of related techniques. To focus on the practicalities, each chapter relates its technical content to one (or maybe two) applications in bioscience or medicine. We hope that the reader should get an appreciation of not only how individual methods can solve real analytical problems, but also the range of applications which can be addressed. The terms image processing, image analysis, and pattern recognition are often used interchangeably. In fact, they refer to different activities, but they overlap and in a given application, it is likely that all three will be used. Chapters 2, 3, and 4 address each of these topics. There is some overlap in their descriptions, where the individual topics merge. The chapters can be read individually, but together they should give a good grounding of the basic methods for dealing with digital images. These are preceded by a chapter on image acquisition. Many users of image analysis acquire their images using a television camera mounted on a microscope. Both of these components are becoming increasingly complex, and an understanding of their properties and limitations separately and in combination is important for achieving the highest quality data. Later chapters deal with more advanced computational methods, in the use of explicit mathematical models of image appearance, and the analysis of three- dimensional data. Not only in microscopy, but also in other research and clinical fields, analysis of structures in three dimensions is of increasing importance. Of course there are omissions. Some readers may feel that we might have included material on confocal microscopy or stereology, for example. These are both large topics which we felt are dealt with very well elsewhere for the practising biologist. We have, however, included the rather less well-known topic of projective stereology, which extends the established stereology literature.Microscope image acquisition? Biological image processing and enhancement Image analysis: quantitative interpretation of chromosome images Pattern recognition: classification of chromosomes Three-dimensional CD) reconstruction from serial sections 3D analysis: registration of biomedical images Model-based methods in analysis of biomedical images Projective stereology in biological microscopy Image warping and spatial data mapping
Чтобы скачать этот файл зарегистрируйтесь и/или войдите на сайт используя форму сверху.