This is the best introduction to image processing published since the late 1970s. The Gonzalez and Wintz book [1] became a classic. Fifteen years later, the Gonzalez and Woods book [2] added, among other things, a wonderful introductory treatment of mathematical morphology for image processing. By 2002, there was published what I consider a third edition [3]. The general topics that were covered never changed; they only got more detailed, and, in fact, better taught. This book can be considered the fourth in this very successful series.

The book contains 12 chapters and three appendices. Chapter 1 is mostly an extended preface. Chapter 2 is one of the best short tutorials on MATLAB I have seen, with an emphasis on how to handle images. General intensity transforms, filtering, and histogram processing are covered in chapter 3. Fourier transforms and frequency domain filtering are discussed at length in chapter 4. Chapter 5 presents a very practical treatment of image restoration, beginning with noise models, and continuing with direct inversion, Wiener filtering, and blind deconvolution. Chapter 6 is devoted to the basics of color image processing, both pseudo-color (as in some medical imaging applications) and full color. The reader should find it very convenient that the color prints appear in the regular text, and are not, as in some books, extracted to a section of glossy pages at the middle or the end. A wonderful, short, and very practical introduction to the use of wavelets in image processing is the theme of chapter 7. Image compression foundation is central to chapter 8, with a discussion of some aspects of Joint Photographic Experts Group (JPEG) 2000.

Mathematical morphology has always been a strong point of the books in this “series,” and chapter 9, where this topic is discussed, lives up to the high standards set by the previous editions. Discussions of edge detection, image segmentation, Hough transforms, and watershed transforms are all collected in chapter 10. Chapter 11 discusses representations, in the form of chain codes, skeletons, Fourier descriptors, and statistical moments. The final chapter presents an overview of pattern recognition methods, based on decision and structural (string matching) processes. The three appendices are mostly concerned with the image processing system developed by the authors, of course in MATLAB, which can be downloaded from the book’s Web site.

The algorithm descriptions and implementation in the text are heavily based on MATLAB, which makes the book perfect for students, though it may be slightly less so for embedded developers who need to implement image processing functions (C code would be more helpful). The strength of the book comes from its very clear and concise presentation of the theory behind the algorithms. The 2002 edition of this book is considered to be the theoretical version, but this one is almost self-contained, and one does not really need to be familiar with the 2002 edition to be at ease with this book.

This very useful and student-friendly book is highly recommended to anyone interested in image processing.