Handbook of Image and Video Processing 2nd Edition (Hardback)

Citation

Connolly, C. (2006), "Handbook of Image and Video Processing 2nd Edition (Hardback)", Sensor Review, Vol. 26 No. 4, pp. 339-339. https://doi.org/10.1108/sr.2006.26.4.339.1

This huge and ambitious textbook aims to supply a reference for researchers and engineers, and an adopted text for students in a range of disciplines. It deals purely with digital images of the real world, and does not cover computer graphics, animation or special effects.

At first sight, the heavy mathematical content is off‐putting, but a more detailed exploration reveals motivational content and material that is both informative and accessible for the beginner in image processing. The introductory chapter is interesting and horizon broadening, revealing the importance of compression techniques in the transmission of television and Internet images. This and the following three chapters on basic image processing techniques make excellent use of pictures, diagrams and graphs to lead the reader carefully through the fundamental concepts such as sampling, quantization, image aliasing, grey‐scale and binary processing. They do include some mathematical formulae, but on cool examination, they turn out to be deceptively simple. At the end of the book are 11 chapters devoted to applications, including fingerprint, face and iris recognition, computer‐assisted microscopy, synthetic aperture radar and computed tomography. The index is also very helpful for the student, allowing cross‐referencing in different parts of the book and giving the meaning of acronyms.

To test the potential as a reference book, I decided to tackle JPEG compression. Going straight to the chapter on lossy JPEG and JPEG2000 standards, I found a clear and straightforward “recipe” for the algorithm that appeared to have all the necessary detail. Whilst there was no worked example, there were web addresses for accessing Java, C and C++ implementations of JPEG2000 Part 1.

Over 130 authors have contributed to the 1372 pages. The high standard of editing makes the book balanced and readable throughout, with nothing glib or shallow. The fingerprint chapter, for example, begins gently with the history and the features inherent in fingerprints, covers the mathematics of feature extraction, classification and matching, and finishes with a brief review of future prospects, giving a very thorough coverage of this application.

Unfortunately, I am not familiar with the 1st edition of this book, but I am impressed by this edition, and satisfied that it is capable of living up to its stated aims. But this hardback weighs in at over 3 kg, so do not try to carry it around!