On January 14, the 2018 general college entrance examination of art in Guangdong Province was held. For the first time this year, the "face recognition system" was used in the unified examination, and all examinees were verified by "face brushing" in the examination room. Compared with the previous fingerprint identification, the process is more convenient, and each candidate only needs 2 to 3 seconds to "cross the machine".
Why face recognition?
Traditional authentication methods include ID card, voice and fingerprint. But the traditional authentication method has many limitations. For example:
1. ID card. The ID card is small, easy to carry, but easy to lose. There are also many cases of fraud on the Internet. Especially since 2012, the railway department has gradually strengthened the verification of identity card and its own identity. For those who need to travel far away, once the ID card is lost or forgotten to carry, it will be very troublesome to buy a ticket and check in the car.
2. Voice. In the past few years, online fraud in the name of borrowing has emerged in an endless stream, not only using other people's social platforms, but also committing crimes by acquaintances. By recording audio and then processing it, voice is sent for fraud.
3. Fingerprints. Compared with ID card and voice, fingerprint is more used to encrypt personal information and check personal data. Just like in recent years, fingerprint verification is needed before the entrance examination to continue the examination. But with the rise of smart phones, fingerprints are also widely used for unlocking and quick payment. But at the same time, there are also illegal transactions of "fingerprint inversion", which makes fingerprint not safe.
Therefore, in order to carry out the verification method with more personal characteristics, the "face recognition" is gradually implemented.
How does face recognition work?
1. Firstly, face information is collected by camera and processed into static or dynamic images;
2. Secondly, through the computer preprocessing system, the collected images are corrected. It mainly includes light compensation, gray-scale transformation, histogram equalization, normalization, geometric correction, filtering and sharpening. In order to eliminate the deviation caused by environment and light;
3. Then, according to the shape description of face organs and the distance between them, the feature data which is helpful for face classification is obtained, and then the face is modeled. The feature components usually include Euclidean distance, curvature and angle between feature points. For example, how many degrees is the angle between a person's glasses and his mouth;
4. Finally, recognition, according to the collected face representation and modeling features, recognize and match the human face, and check the personal information.
What can face recognition system be applied to?
1. Public security criminal investigation
Through face recognition, it will be helpful for the public security organs to compare the target portrait data with the feedback data of key screening places, so as to arrest the fugitives or pre screen the terrorists.
2. Access control system
Using face recognition system in protected areas, such as residential and enterprise, can identify the identity of the person trying to enter conveniently and quickly, and reduce the risk.
By issuing e-passport or ID card, face recognition is carried out in the place where id card or passport needs to be presented, and relevant personal information is automatically retrieved to avoid damage or loss of passport or ID card.
4. Online payment
To avoid fingerprint or password easily to be cracked, through the face recognition system encryption, to prevent non-user use of mobile phones bring adverse consequences.