Accuracy of Fingerprint Biometrics

Accuracy of Fingerprint Biometrics
Page content

How to Read a Fingerprint

There are three ways to read fingerprints: By pattern types, by the size of the patterns, and by their position on the fingers. There is a science to this called dactyloscopy, which is the practice of identifying someone using fingerprints.

The method to identify fingerprints starts from the thumb to the little finger for each hand, then as a group with the four fingers held closer together and the thumbs printed again, individually, for each hand. The fingerprint is a series of ridges and valleys. They are unique, and forensic studies have shown that when you take finger impressions with black ink on a white background paper, the unique outline of the ridges is clear and apparent.

Biometric Fingerprint Devices

The actual fingerprint identification process changes between products and systems. However, the basis of identification is nearly the same. The standard systems have a sensor for scanning a fingerprint, a processor that stores the fingerprint database, and finally a software product, which compares and matches the fingerprint to the predefined database.

421px-Fingerprint scanner in Tel Aviv

Within the database, the process brings together a fingerprint and a reference or PIN number. A comparison to a person’s name or account occurs. In instances of security, the match allows or disallows access, but today something as simple as a time clock or payroll access can also occur.

See Also: Using Biometry for Security and Identification

Accuracy of Fingerprint Biometrics

There are at least three factors that you must take into account to make the biometric accurate.

Human Factor

For fingerprint capture, several elements must be in place. Fingers must be in contact with a sensor surface. As a consequence, human behavior at the moment of biometric capture is a critical issue for the proceeding steps and the accuracy of recognition. An inadequate capture will generate a deficient extraction of biometrical features or even the inability for such extraction. In that regard, to ensure a reliable set of data for assessment, training people in good practices for capture of biometric records is essential.

Device Factors

The technology of biometric sensors also can be a source of errors notwithstanding adequate capture procedures. Environmental conditions and technological properties of devices affect the quality of the captured biometric records. Optical fingerprint devices produce images that are not of high quality; they are very sensitive to light conditions on external and internal environments. This effects the quality of captured images, and processing algorithms must consider a high variation on quality.

Algorithm Factor

A biometric algorithm has three components–enrolling, evaluation, and decision-making. Enrolling analyzes the biometric data, extracts its features, and puts them into a compact biometrical template structure.

Evaluation assigns a similarity score to the comparison between two templates. The first compared record is a query template and the second one the candidate template. If the two record comparisons belong to the same person this is a a genuine comparison. If not, it is an impostor comparison.

The decision-making component uses the score as determined by a matching or ranking algorithm to make a decision about the comparison between two templates. The matching algorithm determines if the comparison is genuine or not. On the other hand, the ranking algorithm uses a predefined integer value to compare the result to the stored database of individuals. It finds the most common templates that match the query template.

Biometric System Diagram


The accuracy of biometric devices depends on human, device, and algorithm factors. Each can provide very accurate assessments, but if any one of them has problems, then this can compromise the accuracy of the biometric device.

See Also: What is a Biometric Reader?


How to Read a Finger Print, Image

Biometric Finger Print Devices Image:

Accuracy of Finger Print Biometrics Image: