Customer Reviews

Feature Extraction & Image Processing for Computer Vision, Second Edition

5 star:		(0)
4 star:		(3)
3 star:		(0)
2 star:		(0)
1 star:		(0)

 

 

How to define and find features in an image is the core of this book. Unlike broader texts on image recognition, like Davies' Machine Vision : Theory, Algorithms, Practicalities, Nixon and Aguado specialise to what researchers might regard as a crucial issue.

There is some amount of background, of course. Like explaining how the Hough transform works in a dual image space. But in this example, as others, the book gives a more indepth mathematical treatment than Davies. So the Hough transform is described for lines, which Davies does, but also for polar representations of lines, and for circles and ellipses. One effect is to give you enough case studies of how Hough transforms are found, that you can make your own, for some geometrical object to search for in your images.

The writing style of this book is very good, and there is a bit of light hearted humour here and there.
I fact I find it rather hard to put down, which is rather rare for a technical book like this.
It's very well pitched for someone like me who has had almost zero experienced at processing pictures with programmes, but who has used a few of these techniques in
software like Corel Photopaint.
The author does a pretty good job at explaining concepts like frequency domain analysis, it could be the first introduction you will ever have.
The explanations on projective geometry are a little too hastily written in my opinion, but I haven't seen an entirely satisfactory text on this subject yet.
The book is a little out of date already as it doesn't mention the SURF technique (which was invented in 2006, see the Wikipedia).
Still, all in all a very good place to start, but by no means a reference book.

Very useful book, full of great informations, accompanied of many examples. Very fast delivery (to Czech Republic), but book came slightly damaged.