5 of 5 people found the following review helpful:
2.0 out of 5 stars
Incomplete and off-topic, March 10, 2008
This review is from: Robot Vision: Video-based Indoor Exploration with Autonomous and Mobile Robots (Hardcover)
This book does not properly address its title or description ("Vision-based Indoor Exploration"). It is a collection of different topics, several of which are unrelated to indoor exploration.
Most of the material is presented as a high-level overview. This might be appropriate for an introductory course, but it is not helpful for a researcher in the field.
Chapter 2 discusses shape recognition for a robot gripper to pick up an object. This does not relate to the book topic, which is autonomous exploration, and it is never mentioned again after this chapter.
There are sections on segmentation and reconstruction in Chapter 5 that use aerial photos as the examples. How does this fit in with "Indoor Exploration"? The examples should use indoor images.
The stated aim in Chapter 1 is to produce a service robot "at a very low price" using cheap cameras. However, in Chapter 5 there is discussion of using a stripe projection system for 3D reconstruction. This is not a cheap system.
Chapter 9 is on OCR, but it is only three pages long including diagrams. Although this relates to the example application in the beginning of the book -- delivering mail to offices -- there is no evidence that the author ever built such a robot. (The photo on the front cover is not the author's robot.)
In Chapter 11, five different algorithms are compared for segmentation. However, the resulting (binary) segmentation images are not shown for any of them. The algorithms include Sobel edge detection and Gabor filtering, which are quite unrelated and intended for different purposes. They might both be used as part of an image processing pipeline, but they are not directly comparable.
It is disappointing to see a book that was published in 2005 with only a handful of references (and several of them self-references) that are later than 2000.
A large number of the references are in German. This is not a criticism, but readers should be aware of this if they do not speak German. Also, the text appears to have been translated from German because it is difficult to understand in places due to the wrong words being used. For example, the "impulse answer" of a filter should be the "impulse response". Confusing technical terms like this make it hard to read. (This is not a criticism of the author, but rather of the poor quality of the technical editing.)
Overall, while it is interesting to read, the book fails to properly address its title.
Help other customers find the most helpful reviews
Was this review helpful to you? Yes
No
