Customer Reviews

Principles of Visual Information Retrieval (Advances in Computer Vision and Pattern Recognition)

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

 

 

As a researcher in multimedia search (at Google), I absolutely love this book! It abounds with tremendously useful information to computer vision and multimedia search developers such as the brilliant chapters on color, texture, shape, and similarity measures. These go far beyond any other books on the market now even though they were written in 2000. I read on the Amazon site some criticisms of this book but I suspect they were written by undergraduates - i.e. No PhD would criticize a work published in early 2001 for only having references from 2000.

My only criticism of this book is that the reader should have expertise in fundamentals of image processing or computer vision beforehand, know what a color space, texture, etc. Read the CV book by Ponce and Forsyth first.

There are only two books which cover the new area
of visual information retrieval. One is by Alberto del Bimbo
and gives a good introduction to the subject. It is
written well, but leaves the reader wondering what the
state of the art is.

This book by Prof. Lew covers the state of the art in the
fundamentals of color, texture, shape, and video. Prof. Lew's
book has each fundamental area written by the leading
expert so the material is fresh and truly state of the art.
These experts include the leading researchers from MIT,
Boston University, Univ. of Illinois, etc.

This is the only choice for an advanced book on visual
information retrieval - one for graduate students or
leading edge developers.

- John

This book was quite a disappointment. The equations the author throws at the reader are extremely abstract and do not get elaborated. I don't think you will be able to implement any of these algorithms in a language like C++ because many details are missing. Most explanations are not self-contained; they usually go along the lines of "[author] uses this function to do blah blah..."

It is indeed true that this is perhaps one of very few books on this important subject. Unfortunately, the content of the book is mainly obselete. It does not contain the start-of-art work since 2000. I am quite disappointed.