Through computer and optical, acoustic, biosensor and biostatistics and other high-tech means, using the inherent physiological characteristics of the human body (such as fingerprints, face, iris, etc.) and behavioral features (such as handwriting, gait, etc.) Identification of personal identity is biometric identification.
It is still very rare to rely on the behavioral features of handwriting and gait to identify people. Personally feel that these two are not suitable for commercial use. Imagine, when paying at the cashier at the mall, let you write a paragraph to the computer to identify it, or the cashier asks you to "take two steps to the machine to see" is a scene?
At present, mainstream identification technologies mainly include fingerprint recognition, voiceprint recognition, vein recognition, iris recognition, and face recognition.
First, fingerprint identification
The fingerprint has lifelong consistency, minimal repetition rate and convenience. By comparing the detailed feature points of different fingerprints, the identification of the customer identity can be realized. For example, on smartphones, company attendance, and security door locks, almost all of them are already using them.
However, fingerprint recognition has limitations. For example, if the finger is slightly damaged or in a wet state, it will affect the recognition accuracy. In addition, sometimes the finger presses the different points of the press, which may cause different degrees of fingerprint deformation and affect the recognition effect.
I just used a fingerprint to boot when I washed my hand, and the screen couldn't be opened for a long time. This kind of thing is often encountered by small friends, and sometimes it is quite anxious.
Second, voiceprint recognition
Voiceprint recognition is a technique based on the development of the biological characteristics of human voice. The voiceprints formed by each person's speech are different, and the identity is identified accordingly.
The obvious shortcoming of voiceprint recognition is that the requirements for the environment are very high, the surrounding environment is too messy and may be unclear; in addition, the voice of the person changes with age, physical condition, mood, etc., which also affects the accuracy of the recognition. degree.
Voiceprint recognition has long been applied to mobile phones, such as when you log in to WeChat. However, according to the face of the brother, there is not much to log in using this method. After all, it is a bit embarrassing to the machine to talk to the machine. If it is used for payment, is it true that each customer is talking about the scene of the machine?
In addition, emotional changes will affect the recognition, the little friends remember to keep the mood stable when facing the machine, otherwise don't blame the machine can't recognize you!
Third, vein recognition
With the vein recognition instrument, the vein distribution map of the individual finger is obtained, and the feature values ​​in the vein distribution map are extracted by an algorithm, and then matched with the feature values ​​stored in the database, thereby realizing identity verification.
Vein recognition is also defective. For example, the vein may change with age or physiological changes, its permanentity has not been confirmed, and the special collection equipment is relatively complicated in design and high in production cost.
Fourth, iris recognition
The iris is an annular part between the black pupil and the white sclera, which contains many detailed features of interlaced spots, filaments, crypts and the like. The iris has formed in the early childhood and will not change.
The recognition rate of iris recognition can reach one-millionth of the level, but at present, the cost of the iris recognition device is too high to be commercialized on a large scale, and the image distortion caused by the lens is also easy to reduce the reliability of the recognition result.
At present, image acquisition devices are relatively large in size and are not flexible to install and use in some places.
Five, face recognition
At present, the face recognition technology has developed rapidly, the recognition accuracy has reached the financial level, the cost of identifying equipment has been greatly reduced, and the market voice and hotspots are also high.
Compared with several other biometric technologies, one of the outstanding advantages of face recognition technology is its non-contact and non-intrusiveness with users. Compare it:
Fingerprint recognition requires the user to press a finger on the sensor;
Voiceprint recognition requires the user to speak loudly;
Iris recognition requires the user to be close to the camera;
Vein recognition requires the user to place the finger and palm on the vein recognition device.
In contrast, face recognition only requires the user to be in the field of view of the camera, and then can be quickly identified, which makes face recognition a user-friendly biometric method.
Before performing face recognition, the face recognition device uses a 3D infrared camera to perform a living body detection on the target by means of a combination of software and hardware to determine whether the collected face is generated by a photo, video or software simulation. The accuracy of the recognition method can reach 99.99%, and the misrecognition rate is only one in a million, so that the identity fraud caused by various face forgery can be effectively avoided. In addition, the current level of face recognition technology is not easily affected by factors such as posture, lighting, and expression.
HEBEI GEMMY IMP.& EXP.CO.,LTD , https://www.hebeiartcandle.com