Face recognition and brush face technology

       In recent years, with the continuous maturity of artificial intelligence technology, face recognition has begun to enter people's lives extensively. More and more scenes are applied to face painting, such as face signing, whether it is going to school or going to work, it is not like before. Need a special punch card machine or manual name, since everyone feels convenient, but also allows the loophole to be checked in to replace the punch card. Because face and fingerprint are the same, there is no forgetting and carrying, and at the same time it is a biometric with high safety factor. Face recognition is often used in payment, security, login and other scenarios that require authorization. The main purpose of these brushing scenes is to confirm the photo of the person in the face and the certificate in the database, which is 1:1 authentication.

       There is also a face-to-face search. This is equivalent to 1:N face recognition, which is to find a target from N's face. It can help us find lost or trafficked children, as long as we have photos of lost children, By comparing with a huge face database, it is possible to find lost children, because these small lost children also need to hold a photo when they enter public again, for example, he goes to elementary school and middle school. Photographs are required for a series of documents such as student ID cards or ID cards. The comparison of artificial intelligence may allow the child's parents or the public security department to identify and find the lost person.

       Face recognition technology is a technique for extracting biometric features of a human face, such as the size of the eye, the length and height of the eyebrows, the distance between the corners of the mouth, the nose, the forehead, etc., and performing authentication based on these features. Face recognition mainly includes face recognition, face feature extraction and face feature matching. Face detection, first determines that the face is detected, can not be the face of the animal, and then further determine the size, position and other information of the face; face feature extraction refers to the location by precisely positioning the key area of the face Point grabbing, to determine that these feature data is exclusive to a person; face feature matching is to determine whether the input face feature data exists or matches with the existing face database, and then finds in the database The face with the highest matching.

       At present, face recognition technology can capture more than 30,000 feature points on the human face, not only can recognize static face images, but also can recognize dynamic facial images while walking or running, including many complicated scenes, such as The night is dimly lit and the face recognition is well recognized. The rate of misunderstanding in our eyes is one in a thousand, and the rate of false recognition of face recognition can be as low as one in a million. Like twins that look very much like each other, it may be difficult to identify them according to the human eye, but it is easy to use face recognition technology.