Customer Reviews

An Introduction to 3D Computer Vision Techniques and Algorithms

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

 

 

Although Hartley's book offers a deeper perspective on stereo 3D reconstruction, it is also aimed at an audience with a higher knowledge on this area, both mathematically and conceptually.
This book on the other hand offers a more simplistic solution but it also offers a good review on disparity and point correspondence estimation, which on Hartley's book is barely covered.
As each of theese books offfers a slightly different point of view on the matter, I suggest having both as good references on the subject.

For professional researchers in computer vision, this book is useful as a summation of the state of the art for 2008. It is comparable to Davies' Machine Vision : Theory, Algorithms, Practicalities (Signal Processing and its Applications) from 2004.

The text starts with a general outline of the problem, giving basic ideas and deploying the maths needed in many of the algorithms. Tensor analysis is widely used and you need to be familiar with it.

As befitting the title, little space is devoted to 2D image recognition. The far more difficult problem is to identify "strong" characteristics of a desired object embedded in a 3D space. Methods like Haar transforms are invoked. But overall, the human eye, backed up by the brain's wetware, is still a more robust detector. The state of the art has further to progress.