Издательство CRC Precc, 2001, -662 pp.Most human activities – including the reading of this text – and interactions with the environment are performed by visual means, more specifically through the capture of electromagnetic waves and subsequent processing by the human visual system. Given the redundancy underlying most that is visual in nature, evolution has produced visual systems that pay great attention to abrupt variations along the objects, such as their boundaries delimiting uniform regions. Indeed, it was very likely the varied geometry of the outlines of uniform objects in the world that led to the human concept of shape and the evolution of visual processing abilities dedicated, or at least related, to the capture, processing, analysis and classification of shapes. Besides, it should be observed that many other natural processes also involve geometric objects, normally interacting through fields and waves. For instance, most biochemical reactions involve the matching between two or more molecules, a process that is governed by their respective shapes and force fields. Similarly, interactions between living beings also involve shape geometry and interactions. For example, the mouth of a whale must be particularly effective for collecting plankton and the wings of birds have to be effective for flying. In brief, the importance of shapes is not only essential to humans, but seems to provide a basic principle underlying most natural processes [Costa and Consularo, 1999]. While the importance of the visual information is now fully acknowledged, defining a scientific area on its own, it was only recently that computer vision came to establish itself as an off-the-shelf tool for both researchers and practitioners. That is to say, it has been the continuous advances in computer technology, including higher processing speed, storage capabilities, better acquisition devices (e.g., cameras and scanners), along with more powerful concepts and algorithms and a progressive cost reduction, that paved the way to the dissemination of imaging techniques through a large number of practical applications in the most diverse areas. Indeed, it is the opinion of the authors of this book that we are undergoing an unprecedented technological opportunity for the effective application of image and shape analysis resources to many areas. Yet, while this tendency has been duly reflected in an ever increasing number of books on imaging, shape processing and analysis, biological and computer vision and pattern recognition, it is felt that only a few (if any) existing textbooks provide an introductory, modern, relatively comprehensive and integrated coverage of shape representation, analysis and recognition, which are closely intertwined. In light of the above reasons, we cannot avoid the commonplace of observing that the current book will fill the gap not only between the areas of shape analysis and classification, including one of the most comprehensive lists of practical shape features, but also between theory and practice. While concentrating on 2D shapes, the current book also provides the basis for the treatment of higher dimensional objects. This book is aimed at serving as a largely self-contained introductory textbook on shape analysis and recognition, including several of the most modern and promising tendencies such as scale-space, analyzing wavelets, fractals, computational geometry and so on. An important trend that has characterized the evolution of shape analysis is its inter- and multidisciplinary nature, involving an increasing variety of concepts from both mathematics and computer science, including differential geometry, Fourier analysis, signal processing, probability and multivariate calculus, to name but a few. This implies that students and researchers alike often experience difficulties while trying to understand many approaches to shape analysis. On the other hand, it is worthy emphasizing that shape analysis is such an important task in so many areas, from biology to material sciences, that it would be highly desirable that experts in all those different areas could understand and apply the techniques explored in this book. As a matter of fact, even those practitioners intending simply to use imaging software should also be acquainted with the basic concepts and techniques in image and shape analysis, in order not only to properly apply the several tools usually bundled into such software, but also interpret the respectively obtained results. As the authors had in mind to create a didactic book that would be accessible to a broad range of readers, a comprehensive and mostly self-contained introduction/review of the involved basic mathematical concepts has been especially prepared and included. In addition, whenever possible the mathematical detail is always preceded by the respective conceptual characterization and discussion and several examples have been included in order to help the assimilation of the presented concepts and techniques.Introduction Basic Mathematical Concepts Shape Acquisition and Processing Shape Concepts Two-Dimensional Shape Representation Shape Characterization Multiscale Shape Characterization Shape Recognition and Classification Epilogue - Future Trends in Shape Analysis and Classification
Чтобы скачать этот файл зарегистрируйтесь и/или войдите на сайт используя форму сверху.