Review: ACM Computing Reviews

Handbook of Iris Recognition
Burge M., Bowyer K., Springer Publishing Company, Incorporated, New York, NY, 2013. 423 pp. Type: Book (978-1-447144-01-4)

Date Reviewed: May 15 2013
Computing Reviews, the leading online review service for computing literature.
Reviewer: Creed Jones Review #: CR141221

The challenge in producing an edited collection of papers is to arrive at a volume that is unified, complete, and consistent. The editors of this book have succeeded. It is an excellent summary of the state of theory, technology, and applications of the important biometric identification modality that uses the characteristics of the iris. John Daugman has been called the father of iris recognition based on his foundational work in the early 1990s. In his foreword to this book, he mentions the critical notion that the accuracy of a biometric system is limited by the statistical properties of the anatomic property used as a basis. He relates iris recognition to the rich characteristics of irises. This is fundamental and necessary, but not sufficient. Imaging and recognition techniques must be devised to accurately use this information. This current volume serves as an excellent presentation of these techniques.

The key overview chapter, “A Survey of Iris Biometrics Research: 2008-2010,” by Bowyer et al., builds on an earlier work that covered the period through 2007. The work covered includes imaging, segmentation, recognition, fusion, sample quality, and other aspects of the overall program. If one could only read a single chapter to become acquainted with the current state of iris recognition, this one would serve very well.

One way to approach the remainder of this book is to think of an iris recognition system from front to back, beginning with an understanding of the physiological attributes and methods for collecting them, then feature extraction and storage, followed by matching and identification, and finally topics related to use and fusion with other biometric modalities. The first level is well represented here. Clark et al. describe “A Theoretical Model for Describing Iris Dynamics,” beginning with tissue dynamic properties and arriving at a (moderately) nonlinear model for iris deformation. Baker et al. discuss “Template Aging in Iris Biometrics,” which is essentially an expression of long-term changes in the iris, showing that it does occur but is likely a slow process.

Ackerman, in “Optics of Iris Imaging Systems,” presents an approachable walk-through of major optical considerations, beginning with the most basic properties of an optical system. Proença discusses the use of light in the visible spectrum (rather than the more common nearinfrared (NIR)) for iris imaging, and asserts that its potential for use in noncooperative scenarios is an advantage. Schmid et al. address the vital issue of “Iris Quality Metrics for Adaptive Authentication,” exploring what quality to require of an iris
image or record and how to deal with exceptions.

Once an iris image is acquired, the accurate and robust segmentation and generation of recognition features is vital. Though the characteristics of the iris are well suited, there are significant methods and techniques involved; several chapters address these methods. Jillela et al. present two papers on the segmentation of the iris itself: a survey of the common methods and a discussion of segmenting problematic images (due to severe lid and lash obscuration, lighting artifacts, and blurring). Klontz and Burge discuss low-quality images as well. Several chapters discuss feature extraction.

In “An Introduction to the IrisCode Theory,” Kong et al. discuss the IrisCode first developed by Daugman and dominant in the marketplace. Bastys et al. propose alternative encodings in “Iris Recognition with Taylor Expansion Features,” and Kumar et al. suggest the “Application of Correlation Filters for Iris Recognition.” Both chapters present good recognition performance. Iris images and template storage are related to feature extraction. Quinn et al. discuss “Standard Iris Storage Formats,” in which “standard” means compliance with the industry standards developed under the International Organization for Standardization (ISO) process.

Several chapters center on system-wide issues of accuracy and use. Phillips and Flynn explain the “Quality and Demographic Investigation of ICE 2006,” the Iris Challenge Evaluation that compared leading systems at that time. Burge and Monaco discuss the fusion of multi-image results in “Multispectral Iris Fusion and Cross-Spectrum Matching,” while Connaughton et al. explore the fusion of faces and iris biometrics collected by a single sensing station. The security of any biometrics system is a valid and popular concern. Venugopalan and Savvides address “spoofing,” where an iris image is reconstructed from an IrisCode template and used to fool an iris recognition system. They see this as an argument for various countermeasures, including template security.

In the interest of testing this excellent volume, I posed several questions about iris recognition and attempted to find the answers within. I was pleased to see that either relevant information was present and quickly locatable, or there were sufficient references for further study. Since the preface claims that this is “the first book to be devoted entirely to iris recognition,” it should be as accurate and complete as a reasonable length allows. It succeeds on both counts. For anyone interested in iris recognition, this book is indispensable.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s