No label defined (Q4023965)

This book is a counterpart ot the author's fundamental treatise ``Ondelettes et opérateurs'' (I: 1990; Zbl 0694.41037, II: 1990; Zbl 0745.42011, III: 1991; Zbl 0745.42012). Whereas the monograph is written for the mathematical analyst, this book emerged from lectures given to a general scientific audience of different specialities. As a consequence the presentation is very informal and mathematical rigor is only secondary. The main goal of this book is the exposition of the intuitive ideas of wavelet theory and their practical applications. Each chapter starts with the description of an important problem chosen mostly from signal and image processing. After the introduction of a fundamental idea the mathematical concepts are made to emerge naturally and almost without effort. The core of each chapter is then devoted to a detailed discussion of these concepts. Particular emphasis is put on the resulting algorithms that intervene in the practical applications. For instance, Chapter 4 deals with image representation and compression. An image contains information on different scales, and going from a fine scale to a coarse one results in a tree structure. Thus it should be possible to calculate an approximation of an image at scale \(2^{j + 1}\) from a representation at scale \(2^ j\) in a natural and efficient fashion. A more technical formulation of this idea leads to the famous pyramid algorithms of image processing and to the mathematical concept of a multiresolution analysis, which are then developed in detail in this chapter. The scheme ``problem -- idea -- mathematical concept -- development of a numerical algorithm'' is executed with great art and persuasion. The book is an excellent model of how an important scientific subject can be communicated to a broader audience. It is very enjoyable to read, and even to the expert it provides new insights and points of view. The book is divided into eleven chapters. Chapter 1 introduces the problems and objectives of signal and image processing, Chapter 2 traces the history of wavelet theory. Chapter 3 and 4 deal with quadrature mirror filters, subband coding, pyramid algorithms, wavelet bases and multiresolution analysis. Chapters 5 to 7 discuss several ideas from time-frequency analysis, such as the Wigner-Ville transform, Malvar bases, and wavelet packets. The remaining chapters give a fascinating account of some more specific applications of wavelet theory in human and computer vision (Marr's program), the wavelet analysis of fractals, wavelets as a tool in the study of turbulence and of distant galaxies. It is significant that wavelet theory is seen as an interdisciplinary field of research with contributions from and implications for mathematics, physics, and signal processing.