I am currently enrolled in a computer vision class and we used the book to write a lot of the code for image processing. So far we have found a lot of typos and bugs in the code which wasted a lot of time in getting this to work. Some like an incorrectly calculated quadratic formula where easy to see in Major Axis but others like errors in the Moments code took a lot of debugging to find. Would love a second revised edition with the code thoroughly checked for errors. Oh and it turns out that the Major Axis code does not even work correctly, allowing the major and minor axis to switch after 90 degrees of rotation, this is something you probably wouldn't notice using it to process a static image but you can see when working with live video. I understand that this book is meant for people already in the field but that is not an excuse for writing bad code.

Potential buyers, please keep in mind that this book is for people that already know image processing theory or are enrolled in a course. We wrote this book for the student engineer to rapidly be able to code algorithms. They are very simply written, more so for understanding and ease of typing (most are one small page) than for algorithmic elegance or efficiency. Note that the book is very small and has a soft leatherette cover--this is so it fits comfortably in your back pocket and can be pulled out as needed. There is no CD, type the code in yourself and learn as you go. Some things are done for a reason, enjoy!

Total absence of theory. Algorithms very poorly documented. However, it does open the eyes to the available algorithms in image processing. If you want to know how they really work, look elsewhere. At > $30, this book is an utter dissaapointment. At $3., it would have earned 4 stars. The least the authors could do is publish a CD to go along.

It intruduced broad terms of image processing tech. It's practical and easy to understand. It might only take you a couple of days to finish reading it.

Very useful book. It provides *many* useful algorithms. The book is tiny, 3x5x1 so pound for pound it is a great (and portable!) book.

Some issues however detract from a 5 star rating:

a) The code examples are lame (lots of single letter variables) and poor C idioms, and some silly extra calculations, like counting pixels one at time in a loop when a single multiply would yield the same result.

b)The book feels old even though it was written in 1993. It has a codec for MacPaint files. If this doesn't ring a bell to you, MacPaint was written circa 1985 and it supports two bit images only. Most algorithms are tweaked for the B/W and/or grayscale versions so the reader will have to interpolate/derive her own color versions

C) The descriptions of the algorithms are suboptimal, but heck it is still orders of magnitude more useful than one of the fat texts on the subject. Coupled with the code examples you can easily figure out any non-clear descriptions without having to wade through a giant tome full of pseudo code.

Would I get it again - absolutely. I even made a special place for it (the top of my monitor so it won't get lost).

This book presents many algorithms and their C implementations in alphabetical order. Many of the entries show the result of the algorithm's application on an image. A table of contents groups the algorithms into classes, such as image fundamentals (convolution, correlation), morphological filters, and nonlinear filters. The basic stuff is all there. Also included in the appendix are programs to read and write image data, which can be used in conjunction with the C code listed for each algorithm for a working solution.

The theory is not emphasized, so readers should look elsewhere for that information. However, considering its small size, this book offers a lot of very useful information.