Introduction to several common biometric technologies

Analysis of several common biometric technologies, including fingerprint, palm shape, finger vein, retina, iris, facial recognition technology ...

fingerprint

Fingerprints refer to the strip lines on human fingers. Their formation depends on the environment during embryonic development. At present, many countries have stored fingerprints as official collected personal information data in the identification database, which has become a means of identification by the judiciary.

As the most traditional and mature biological identification technology, fingerprints have the following two outstanding advantages: stability, from the complete formation of the fetus at 6 months to the decay of the body after death, the type, structure and statistical characteristics of the fingerprint There has been no obvious change in the overall distribution, etc .; uniqueness, so far, two people with identical fingerprints have not been found.

Fingerprint comparison usually uses the feature point method to extract the divergence of the convex curve on the fingerprint or the cut-off part (end point) in the fingerprint to identify. In recent years, with the development of electronic technology, the performance of automatic fingerprint recognition has been greatly improved. Generally, recognition generally takes only about 1 second, the error rejection rate is less than 2%, and the error acceptance rate is less than 0.0001%.

However, in order to improve reliability, the fingerprint recognition device has certain requirements on the placement of the opponent. In addition, the system also has certain requirements on the actual situation of the fingers, and the recognition rate of dirty or dry fingers will be reduced.

For the end user, fingerprints have been entered into the official identity database as personal information in many places and used in the access control system. If the management is not good, there is a risk of leakage of personal information.

Palm shape

Palm recognition is a technology that emerged in the 1980s. The device for palm recognition is based on three-dimensional measurement of the geometric shape of the human hand. Because each person's hand shape is different, it can be used as a condition for recognition. The identification function is mainly realized by determining several features of the human hand, including the length, width, thickness and surface area of ​​the palm and the width of the finger in different parts, the length of the finger, the thickness of the finger and the curvature of the curved part of the finger. These data provide a unique combination of relationships, ensuring fast, accurate and reliable identification.

Normally, the error rejection rate of palm recognition devices is about 0.03%, and the error acceptance rate is 0.1%. The system only needs 1 second to complete a recognition.

Finger vein

Finger vein recognition is a technique that uses near infrared rays to illuminate the veins of a living human finger, acquires vein images, and encodes and authenticates them. Medical research has proved that the shape of human finger veins is unique and stable, that is, each person's finger vein image is different, and the same person's finger vein image is also different; healthy adult vein shape no longer changes, This provides a medical basis for finger vein recognition. Compared with fingerprint recognition technology, it is hidden inside the body and requires living finger veins, so the chance of being copied or misappropriated is very small, the psychological resistance of the examinee is low, and it is less affected by physiological and environmental factors, which overcomes dry skin , Oil stains, dust, skin surface abnormalities and other factors, the original vein image is captured to digital processing, the whole process is less than 1 second, and has a triggerable high accuracy recognition rate (the incidence of false rejection is about 0.01%, false acceptance occurs The rate is about 0.0001%) and many other important features, making it far superior to fingerprint recognition technology in high security and ease of use.

Retina

The principle of retina recognition is to distinguish each person by analyzing the blood vessel pattern on the retina. Retinal scanning is to illuminate the retina with low-intensity infrared rays to take an image of the main blood vessels. Because the retina is located behind the eyeball, the user needs a high degree of cooperation during the collection process to ensure correct illumination and alignment of the retina, and requires to stand 2-3 feet and stay still for 1-2 seconds. More importantly, the identified person's worries about retinal scanning technology, which may affect eye health. For these reasons, retinal recognition technology has not become the mainstream technology in biometric technology.

Iris

The iris is a ring-shaped color tissue around the pupil. It has rich and varied texture patterns, which form the basis of iris recognition. The formation of the iris is determined by genetic genes, and the expression of human genes determines the morphology, physiology, color and overall appearance of the iris. By the age of about two, the iris has basically grown to a sufficient size and entered a relatively stable period. In addition to the unusual changes in the appearance of the iris caused by unusual abnormalities, physical or mental trauma, the shape of the iris can remain unchanged for decades. In addition, the transparent cornea on the outside of the iris connects it to the outside Isolation, therefore, a fully developed iris is less susceptible to changes from outside damage. Iris recognition technology is to scan and image the iris pattern through a kind of light that is similar to infrared rays, and determine the degree of similarity through the exclusive OR operation of the pattern pixel bit. The iris recognition process first needs to separate the iris from the eye image, and then perform feature analysis. In theory, the probability of finding two identical irises is one in 1.2 million, which is also the most accurate of all known biometric technologies. of.

Iris recognition has not been widely used in the civilian market because of its complicated equipment, short scanning distance, and users' psychological concerns about health.

Facial

Facial recognition is a kind of biometrics technology for identity recognition based on human facial feature information. Use the camera to collect images or video streams containing faces, and automatically detect and track faces in the images, and then take a series of related technical measures for the detected face images, including face image acquisition, face positioning, and face Recognition preprocessing, memory storage and comparison identification, to achieve the purpose of identifying different people.

The facial recognition device mainly analyzes the shape and features of the face, and these features include the contours, shapes, and position relationships of the eyes, nose, mouth, eyebrows, and face. Facial recognition is very complicated due to the different brightness and angle and facial expressions. In addition, the facial recognition system is very sensitive to direct light, especially sunlight. In this case, the false rejection rate will greatly increase, and in some cases, it may not even be possible to recognize.