The basic principle and process of fingerprint identification technology-biometric identification technology

The basic principle and process of fingerprint identification technology

Although fingerprint identification technology has entered the civil field, but its working principle is still relatively complex. Unlike manual processing, biometrics companies do not store fingerprint images directly. Over the years, the biometric companies and their research institutions have studied many fingerprint identification algorithms (the United States law that the fingerprint image is personal privacy, so can not directly store fingerprint images). But all kinds of recognition algorithms ultimately come down to finding and matching fingerprints on fingerprint images. This is the basic principle of fingerprint identification technology, that is, fingerprint image acquisition and fingerprint characteristics.

The characteristics of the fingerprint

In a general sense, two types of fingerprints can be defined to verify the fingerprint: The general feature and the local feature.
The overall characteristics are those that the eye can observe directly. It includes:

1, basic pattern

Common fingerprint patterns are ring, bow, spiral, and other fingerprint patterns are based on these three basic patterns, but a rough classification, only rely on pattern types to distinguish fingerprints is not enough, but through the classification can be more convenient in the large database to search for fingerprints.

2, Mode area

The pattern area is the area that contains the pattern feature, that is, it is possible to distinguish the fingerprint from the pattern area.

3, Core point

The core points are in the progressive center of the fingerprint lines, which are used to read fingerprints and reference points when compared to fingerprints.

4, Triangle Point (Delta)

A triangular point is located at the first fork point or breakpoint from the core point, or two lines converge, isolated, equipped, or point to these singularities. The triangle points provide the counting of fingerprint lines and the beginning of tracking.

5, model line (Type Lines)

A model line is a cross in which the pattern line begins to run parallel to the line in the fingerprint surround mode area, which is usually interrupted at short notice, but its outer line begins to extend continuously.

6, the number of lines (Ridge count)

Number is the number of fingerprint lines in the pattern area. In calculating the number of fingerprints, usually first connect the core and triangle points, this line and the number of fingerprint lines can be considered to be the number of fingerprints.
Local characteristics refer to the nodes on the fingerprint. Two fingerprints often have the same overall characteristics, but their local characteristics cannot be exactly the same. 

The local feature points have the following types:
1) node: Fingerprint lines are not continuous, smooth and straight, but often interrupted, forked or bent. These breakpoints, bifurcations, and turning points are called "nodes." Is that these nodes provide confirmation of the fingerprint uniqueness.
2) Endpoint: The place where a stripe line terminates;
3) Bifurcation point: a stripe line split into two places;
4) Center point: The geometric center of the fingerprint, which is the place where the maximum curvature of the line is produced;
5) triangle point: Three kinds of different directions of the lines converge place;
6) Cross: The place where the two stripes line produce intersection;
7) Small island: A very short line of stripes;
8) Sweat glands hole: the small hole on the ridge line, the Department of sweat glands. 

These features are used to distinguish between different fingerprints. Where:

1. Endpoints and bifurcation points are the most commonly used features. The usual algorithms have to keep track of their location and direction. 

2. Central and triangular points are widely used in criminal investigation systems and are not commonly used in civil systems. Because the collector used in these applications tends to be smaller, it is difficult to collect the central point and the triangle point completely. 

3. Intersections and islets are often not used in practical systems because of computational difficulties. 

4. It has been suggested that the use of sweat glands for fingerprint identification, but this method requires a fingerprint acquisition device to have a very high resolution. So it's not used in the actual system.

The process of fingerprint identification

Fingerprint identification consists of two processes, namely, registration process and identification process. The principle is as shown in Figure 1

Fig. 1 The process of fingerprint identification

In the registration process, the user needs to collect fingerprints first, then the computer system will automatically feature extraction, the extracted features will be stored as a template in the database or other designated places. In the identification or verification phase, the user first needs to collect the fingerprint, then the system automatically carries on the Fingerprint Library template alignment, then gives the comparison result.

In many cases, users may also want to enter some other auxiliary information to help the system to match, such as account number, user name and so on. This process is a common process that applies to all biometric recognition techniques.

The way of fingerprint collection

At present, there are three kinds of fingerprint collection devices commonly used in the market: optical type, silicon chip type and ultrasonic type. The performance comparisons of these three types of acquisition techniques are shown in table 1:

Table 1 Performance comparison of optical, capacitance and ultrasonic sampling techniques

1,Optical type

The optical fingerprint collector is the earliest fingerprint collector, and it is the most popular one at present. It has the following advantages: 

1) The longest use, has withstood the actual use of the test, 

2) temperature and other environmental factors to adapt to the good;

3）prices are relatively low, high resolution, can reach more than DPI.

At present, has also appeared the grating lens replaces the prism and the lens system's collector, the photoelectric conversion CCD device also already replaced the CMOS imaging device, thus omitted the image capture card, obtains the digital image directly.

2, Silicon chip type

Silicon chip fingerprint collector appeared in the late 90, most of the silicon chip measurement is the finger surface and the chip surface of the DC capacitor field, the capacitor field by a/D conversion into gray-scale digital image.

3, Ultrasonic type

The ultrasonic fingerprint collector may be the most accurate fingerprint collector, but it is not mature yet and has not been applied in large scale.