Customer Reviews

Multiple View Geometry in Computer Vision

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

 

 

This book is very complete and rigorous in its explanations of the theory. However, I just think I like the approach in An Invitation to 3-D Vision a bit better. This book is better illustrated than that one and is more careful in its explanations, but this book just seems more focused on providing complete proofs than giving you a feel for how you would approach a real problem. Even the exercises are more along the lines of proofs. I like how An Invitation to 3-D Vision ends the book with a complete example. In all fairness, though, this book does have quite a bit of Matlab code on its website.

The book begins with some background material on 2D and 3D geometry. Then the author explains single-view geometry and how cameras map an image in 3D space to an image. Two-view geometry is next, with the author describing the epipolar geometry of two cameras ahd projective reconstruction from resulting image map correspondences. Part three of the book extends ideas to three cameras and the resulting trifocal geometry. The final section of the book takes the algorithms of the book to N views. Thus this book has a simple and straightforward structure that belies the complexity of the material.

If you are really researching this subject you should probably have this book for explanation, illustrations, and rigor, and the Invitation book for enlightenment through a good example-based approach. You should also have Introductory Techniques for 3-D Computer Vision as a text on the individual pieces of algorithms involved in 3D vision. And don't even think about getting into this subject unless you already have a firm foundation in linear algebra, image processing, and computer vision in general as found in Computer Vision, which is my favorite introductory computer vision text.

It is the best book in this area that I have seen up to now. It is well-organized and all the notations and words are friendly to beginners and even experts in this field. Included materials are really tracing the latest advanced techniques. Actually, it is great that there are a lot of exercises at the ends of each chapters but there is no sufficient solutions or detail explanations to each questions.

The first edition of this book could have been much better written. It took up a lot of topics, but treated each in a summary fashion. In fairness, though, I must say that this may be as good as any other book with its aim and scope, and better than some. Any writer on computer vision faces the problem of guessing who the reader is likely to be and what the reader's background is. Also, each of the various topics really merits a sizable book. In particular, the mathematics needs a truly mathematical treatment in a separate book. I have not seen this second edition, but there was room for improvement over the first edition.

Great book in this subject. The good things are...
- Clear theoritical introduction which many books miss to add
- Great appendixes about some mathematical theories necessary to understand this book
- Great organizaion of book chapters and coherent topics for each
- Bonus: Chapters for estimating hemographies and the math behind that
I cannot see any cons of the boox except that there is no clear road map to go into specific topic. For example, I am interested in multiple view geometry only (Trifocal tensor and above), I cannot figure out what I should read and what I can skip. I have to figure my way through.
After all, the best in the subject. Recommended.

I find the book very useful, it is full of practically useful content. Formulas, theorems, lots of examples and illustrations. Overall very easy to read and understand, though requires you to recall your forgotten mathematical skills. The book does present what it claims on the first pages, so read the abstract and judge for yourself if you need the book. For my purposes, I found it to contain all the material I needed to perform certain image photo transformations and compositions. There is also lots of reference material, in terms definitions, formulas and theorems with proofs. And it's good to have it all in one place.

Overall I would say it is worth the money.

The book succeeds in introducing you to the world of multiple view geometry. Specially the math and geometry concepts associated with it. In my research, I had to work on stereo images and this book provided very good information about it. The algorithms are presented very clearly and have been easy to implement (at least in Matlab).

It's a good reference book to have.

If you're interested in photogrammetry, buy this book. Extremely well written, extremely informative, and more clear than I could have hoped for. The only thing it doesn't provide is written out code (not even pseudocode) - just a plain-English description of each algorithm's steps.

My lab has the first edition of this book. Everyone likes it. That's why we order a second book. I have not read through the second edition yet, but this book rocks!

The book has a lot of valuable information for those who are working in computer vision. The book however is fairly terse on many subject and requires careful reading.

I received the fourth printing a few weeks ago. It is missing pages 177-208. That includes all of chapter seven, on camera calibration. Ridiculous.