Computing Reviews
Today's Issue Hot Topics Search Browse Recommended My Account Log In
Review Help
Search
Registration and recognition in images and videos
Cipolla R., Battiato S., Farinella G., Springer Publishing Company, Incorporated, Berlin, Germany, 2013. 400 pp. Type: Book (978-3-642449-06-2)
Date Reviewed: May 15 2014

In recent years, computer vision has emerged as one of the most popular areas in computer science, primarily because of significant advancements in the capabilities of hardware and because of improvements in the efficiency and effectiveness of algorithms. This book provides an in-depth overview of the recent advances in computer vision, with a particular focus on images and videos. It is an edited volume that compiles the outcomes of the now-annual International Computer Vision Summer School (ICVSS).

The book assumes that the reader already has a prior working understanding of the important concepts in computer vision. Thus, the book seems to be ideal for students, researchers, and anyone already working in the area of computer vision or interested in learning the intrinsic details of the algorithms used in computer vision.

The book’s 11 chapters cover the human visual field, feature extraction and description, feature matching and image registration, object detection and recognition, object tracking, and image segmentation. Each chapter contains important references to literature.

Of particular interest is the chapter on “Descriptor Learning for Omnidirectional Image Matching.” The authors address the complex problem of describing invariant features in omnidirectional images. Traditional techniques such as scale-invariant feature transform (SIFT), speeded up robust features (SURF), and the like rely on the assumption that a keypoint feature appearing captured from two different viewpoints will have the same descriptor. Although the assumption holds in most cases, in the special case of omnidirectional imagery this does not produce very accurate results, so the design of invariant descriptors is very complicated. The authors suggest the reformulation of the problem as a learning problem where descriptor invariance is learned using a neural network from training sets of similar and dissimilar descriptor pairs. Their method seems to outperform other approaches.

Another noteworthy chapter is “Socially-Driven Computer Vision for Group Behavior Analysis.” The authors, Cristani and Murino, try to answer the question “What is a group of people?” in a video by incorporating principles from social, affective, and psychological disciplines to the available computer vision techniques. The chapter provides an extensive overview of how social and psychological information can be incorporated in the formulation of the solution, and the authors present several preliminary results that validate their claims. Although this work is still in the early stages, the reported results seem quite promising and seem to indicate that a cross-disciplinary solution is possible.

Overall, this is a very good book and I would recommend it to anyone interested in learning more about the current state-of-the-art computer vision methodologies. This book is definitely not for everyone--it requires a strong math background as well as a prior working understanding of the basic concepts of computer vision. In my opinion, the book will be of interest to the wider computer vision community.

Reviewer:  Charalambos Poullis Review #: CR142284 (1408-0633)
Bookmark and Share
  Reviewer Selected
 
 
Scene Analysis (I.4.8 )
 
 
Video (H.5.1 ... )
 
 
Applications (I.4.9 )
 
 
Multimedia Information Systems (H.5.1 )
 
Would you recommend this review?
yes
no
Other reviews under "Scene Analysis": Date
Connected components in binary images: the detection problem
Ronse C., Devijver P., John Wiley & Sons, Inc., New York, NY, 1984. Type: Book (9789780863800160)
Aug 1 1985
Visual reconstruction
Blake A. (ed), Zisserman A. (ed), MIT Press, Cambridge, MA, 1987. Type: Book (9789780262022712)
Jun 1 1988
Experiments in computing optical flow with the gradient-based, multiconstraint method
Mitiche A., Wang Y., Aggarwal J. Pattern Recognition 20(2): 173-179, 1987. Type: Article
Jul 1 1988
more...

E-Mail This Printer-Friendly
Send Your Comments
Contact Us
Reproduction in whole or in part without permission is prohibited.   Copyright 1999-2024 ThinkLoud®
Terms of Use
| Privacy Policy