Automatic Fingerprint Recognition Systems

Citation

Andrew, A.M. (2005), "Automatic Fingerprint Recognition Systems", Kybernetes, Vol. 34 No. 6, pp. 902-902. https://doi.org/10.1108/03684920510595599

The 21 chapters are from different leading workers in the field, invited by the editors and not associated with a formal conference. Almost all of the contributors have affiliations in USA, though Canada, Italy and Japan are also represented. The topic is of current interest not only because of its forensic significance but also because it contributes to the “hot topic” of biometric techniques for identity checking to protect against terrorism and fraud.

The editors review the contents in their preface, from which the following is abstracted.

Chapter 1 summarises the history of fingerprint recognition and its social significance. In chapter 2, new developments in inkless sensing are presented with specific emphasis on radio frequency imaging. Images may require enhancement and correction of distortion and the next two chapters describe ways of assessing image quality and hence the need for improvement. Then chapters 4‐6 describe the means of automatic enhancement using statistical methods and Fourier and Gabor filters.

In chapter 8 the adaptation of established manual methods to automatic recognition is considered, including the recognition of minutiae and broad classification by ridge patterns. In chapters 9 and 10, other approaches are described depending on the Karhunen‐Loève transform and on hidden Markov models.

Subsequent chapters deal with fingerprint authentication and matching, with reference to distortion‐tolerant filters and eigenfeatures, and higher‐level issues including recognition reliability and methods to be applied when images from two or more fingers of a subject are available.

Forensic aspects arise again in two chapters dealing with treatment of latent images, with comparison of automatic and manual methods. A means of generating synthetic fingerprints for testing recognition schemes is described, as well as the means of storing and transmitting fingerprint data, the latter with attention to security as well as fidelity.

This is an admirable review of the state‐of‐the‐art in this important field, and many of the principles are likely to be applicable to other methods of biometry, as well as to the study of pattern recognition in general, including perception in living systems. Each chapter has a substantial bibliography and there is an overall subject index.