Textbook in PDF format

Several disciplines dealing with “the technology of digital images” emerged over the past few decades and are taught at universities worldwide. Image Processing is about algorithms that filter, enhance, and restore images: this is an “image-idimage-out” technology. Coding and compression, and image reconstruction from projections are topics often incorporated in digital image processing books and courses. Computer Graphics deals with image synthesis: the goal is to simulate the physical image generation process and produce credible glimpses of “ghosts” residing in computers’ databases, figments of the programmer’s imagination or results of complicated numerical calculations. Computer Vision, a most frustrating discipline, addresses issues of image analysis and interpretation: the general aim is to accomplish “image-iddescription-out.” This endeavor leads to very difficult, often ill-posed and not well defined, inverse problems. But these issues are so interesting that recently it has become fashionable among famous mathematicians and physicists to move into this xciting new area of research. Imaging traditionally refers to the process of producing images from data. The data can be gathered in various ways, either by conventional cameras detecting the flux of photons falling on arrays of sensors, or by reading the echoes of waves returning from a medium whose space-varying properties we want to map. Professor Ronald Bracewell, whose wonderful textbook on Fourier transforms should be on every engineer’s bookshelf, sets out to provide in Two-Dimensional Imaging a solid theoretical basis for the student about to embark on the exciting adventure of dealing with images on the computer. He does this brilliantly, his writing is entertaining, his points are clever and clearly made. “Every book is autobiographical,” someone remarked, and this is quite true of Professor Bracewell’s recent book as well. He shares with his reader an expanded and organized story of his own frequent forays into the emerging technology of imaging. Coming from Ronald Bracewell, an outstanding astrophysicist and experienced teacher, this was bound to be, and indeed is, quite an opinionated and exciting story. Feynman remarked in his celebrated Lectures on Physics, when devoting a chapter to the topic of color vision, that “the separation of fields is merely a human convenience, and an unnatural thing.” “Nature,” he said, “is not interested in our separations, and many of the interesting phenomena bridge the gap between the fields.” Professor Bracewell knows this and covers a truly amazing variety of topics connected to the science of digitized images with little regard for the conventional boundaries that separate the various types of image-related endeavors. In his book, image acquisition, processing, synthesis, and analysis topics mix and match, and relevant subjects, not commonly discussed in texts on digital images, such as imaging as a deconvolution process, diffraction theory of sensors and radiators, and aperture synthesis, and interferometry, are given serious consideration. At each point in the book, well-chosen and interesting examples and applications of the theoretical tools are discussed. The classical subjects of image representations in the frequency domain (via Fourier and Hartley transforms, of course!), filtering and restoration, are treated with an emphasis on analyzing the relationship between the continuous-domain theory and discrete domain algorithms, an absolutely necessary prerequisite when dealing with images as pixel arrays