There are many ways to realize fingerprint identification. Some imitate the traditional methods of comparing local details of fingerprints by public security departments; Some are directly identified by all features; Others use more unique methods, such as the ripple edge pattern of fingerprints and ultrasound. Some devices can measure finger fingerprints in real time, while others cannot. Among all biometric technologies, fingerprint identification is the most widely used one at present.
Fingerprint identification is more suitable for indoor security system, because it can provide sufficient conditions for users to explain and train, and the operating environment of the system is controllable. Because of its relatively low price, small size (it can be easily integrated into the keyboard) and easy integration, almost all applications in the workstation security access control system are fingerprint identification.
Geometric recognition of palm
Geometric recognition of the palm is realized by measuring the physical characteristics of the user's palm and fingers, and advanced products can also recognize three-dimensional images. As a mature method, geometric recognition of palm is not only good in performance, but also convenient to use. Suitable for a large number of users, or users do not often use it, but it is easy to accept when using it. If necessary, the accuracy of this technology can be very high, and at the same time, the performance of biometric technology can be flexibly adjusted to meet a wide range of use needs. Palmprint reader is widely used and easily integrated into other systems, so it has become the first choice technology for many biometric projects.
voice recognition
Speech recognition is a kind of recognition technology by analyzing the physical characteristics of user's voice. Although some speech recognition products have entered the market now, they are not convenient to use, mainly because of the great variability of sensors and voices. In addition, compared with other biometric technologies, it uses more complicated steps and is more inconvenient in some occasions. A lot of research work is under way, and we believe that speech recognition technology will make great progress.
Retinal recognition
Retinal recognition uses a low-intensity light source emitted by an optical device to scan a unique pattern on the retina. There is evidence that retinal scanning is very accurate, but it requires users to look at the receiver and stare at a point. This is inconvenient for people who wear glasses, close to the receiver and uncomfortable. So although the retina recognition technology itself is very good, the user's acceptance is very low. Therefore, this kind of product is still a non-mainstream biometric product, although it was redesigned in the 1990s to enhance connectivity and improve the user interface.
Iris recognition
Iris recognition is a technology that causes less interference to people in biometric identification related to eyes. It uses quite common camera elements and does not require users to touch the machine. In addition, it can achieve higher template matching performance. Therefore, it attracts the attention of all kinds of people. In the past, iris scanning equipment had no advantages in simplicity of operation and system integration, and we hope that the new products can be improved in these aspects.
Signature recognition
Signature identification has advantages over other biometric technologies in application. People are used to signing as a way to confirm their identity in transactions, and its further development will not make people feel much different. Practice has proved that signature recognition is quite accurate, so the signature can easily become an acceptable identifier. However, compared with other biometric products, there are few such products now.
Gene recognition
The first step in making this genetic ID card is to obtain relevant genes, test them, select characteristic loci (DNA fingerprints), and then load them into the central computer repository, thus making a genetic ID card. If you like to add personal medical records and carry out genetic testing, it is also possible. After the genetic ID card is issued, doctors and related medical institutions can use smart card readers to read the medical records of relevant personnel.
Gene recognition is an advanced biometric technology, but due to technical reasons, it can not be sampled in real time and recognized quickly, which limits its wide application to some extent.
In addition to the biometric technology mentioned above, there are also technologies for identification through characteristics such as smell and earlobe. But they can't walk into daily life yet.
Vein recognition
Vein identification, which uses near infrared ray to read the vein pattern and compare it with the stored vein pattern to carry out personal identification technology. Its working principle is that the blood flowing in human fingers can absorb light with a specific wavelength, and a clear image of finger veins can be obtained by irradiating fingers with light with a specific wavelength. Using this inherent scientific feature, the obtained image is analyzed and processed to obtain the biological characteristics of finger vein, and then the obtained finger vein characteristic information is compared with the pre-registered finger vein characteristics to confirm the identity of the registrant.
The vein recognition system first obtains the personal vein distribution map through the vein recognition instrument, extracts the characteristic values from the vein distribution map according to a special comparison algorithm, obtains the images of fingers, palms and dorsal veins through the infrared CCD camera, and stores the digital images of veins in the computer system, and stores the characteristic values. During vein comparison, the vein map is taken in real time to extract feature values, and features are extracted from digital images by advanced filtering, image binarization, thinning and other methods, which are compared with vein feature values stored in the host computer, and vein features are matched by complex matching algorithms, so as to identify individuals and confirm their identities.
Gait recognition
Gait recognition, which uses the camera to collect the image sequence of human walking process, compares it with the stored data after processing, and achieves the purpose of identity recognition. The Institute of Automation of China Academy of Sciences has conducted some research. Gait recognition, as a biometric identification technology, has unique advantages that other biometric identification technologies do not have, that is, the recognition potential in the case of long distance or low video quality, and gait is difficult to hide or disguise. Gait recognition is mainly to analyze and process the moving image sequence containing people, which usually includes three stages: motion detection, feature extraction and processing, recognition and classification.
Electronic character recognition
Also known as face recognition, or portrait recognition, it uses advanced biometric technology in the field of artificial intelligence, especially computer technology that uses people's visual feature information to identify themselves. Overview of Person Recognition In fact, person recognition in a broad sense includes a series of related technologies to build a person recognition system, including person image acquisition, person location, person recognition preprocessing, identity confirmation and identity search. In a narrow sense, identity recognition refers to the technology or system of identity confirmation or identity search through people.