Many application packages and programming languages provide some graphic capabilities to view results of data processing. Often enough, this is for standard needs, but data visualization is something more, and Telea’s book proves it: data visualization is a field of its own, with its own problems and challenges.
In a compact yet informative format, the author introduces us to this discipline from fundamentals to some frontier techniques. The book revolves around the concept of a visualization pipeline or the notion that data visualization begins long before manipulating some graphic function in any given language; data visualization begins at initial data gathering. There are plenty of smart ways to do this, as well as ways that give rise to data inconsistency. Final visualization on a computer screen is only the last step of many, and the book details them all in a thorough and competent manner.
The book is divided into 12 chapters and four appendices. The first chapter is a rapid overview of the data visualization process. Chapter 2 is about data representation: continuous and discrete data, data sampling techniques, and the building of datasets that are the original data in a form ready to be processed by a computer algorithm. Chapter 3 is the most important, for the reasons mentioned above, in that it details the visualization pipeline and its four steps: data importing, data filtering and reconstruction, data mapping, and data rendering. This chapter also explains that a good image, resulting from a good overall process, has two mathematically defined properties: it is invertible--or it conveys the exact meaning of the data--and it is unambiguous--or it conveys only one meaning. Subsequent chapters deal with problems, techniques, and solutions to visualize: scalar, vector, and tensor quantities; domain modeling techniques that improve image quality by modifying the sampling domain rather than the sampled data; image visualization--how to process digital images and extract images from video sequences; and volume visualization--how to build the image of whole three-dimensional (3D) scalar volumes on a flat surface. The last chapter introduces information visualization, a fascinating and quickly evolving branch of the field that allows the visualization of abstract data types such as graphs, database tables, and software versions.
The appendices are surveys of popular visualization software packages divided by application domain (scientific visualization, medical imaging, and information visualization) and described in terms of their features.
As the author is a professor of scientific visualization and computer graphics at the University of Groningen, the Netherlands, this book has all the defects and qualities of academic work. It is certainly a dense book that requires concentration and keeps the casual reader at a good distance; on the other hand, the text, examples, figures, equations, and code fragments are very well interwoven, making for terse, even pleasant reading; even the editorial hardcover format conveys a sense of rigorousness and reliability.
The book could make an excellent textbook in a data visualization course, and was probably born out of such a situation, but it will also be useful to information and communication technology (ICT) practitioners who wish to stay current on the latest technological developments.
