Face recognition technology is not complicated. The front-end device is responsible for face capture and certificate information reading. The read information is uploaded to the cloud server and other processing systems for comparison, and the server feeds back the comparison results to the front-end device for display. This process is roughly divided into three steps:
The first step is face detection, using the camera to collect the face image files of the people's faces or take their photos to form face image files, and then encode and store these face image files to establish files;
The second step is face analysis, which distinguishes the contours of the eyes, mouth, nose and other different regions in the face;
The third step is face recognition, which encodes with the current facial pattern Compared with the file inventory, through the comparison, to determine whether it is the same face.
At present, face recognition technology is mainly used in finance, security and its surrounding areas, and is expected to be popularized to all aspects of daily life in the future. However, many people worry about the accuracy and anti risk ability of face recognition. For example, can face recognition be achieved after makeup and plastic surgery?
Face recognition technology has reached the industrial and large-scale application standards. In the third step of the "three steps" of face recognition, there is an important technology called key point capture, which is to accurately recognize the face through the whole sampling of many sampling points of the face. Therefore, the influence of simple shape and makeup can be ignored, and the recognition ability of the machine system has exceeded that of human beings. Whether face recognition after plastic surgery can be realized depends on the change of facial features. If it's just cosmetic surgery, such as raising the bridge of nose, shrinking chin or cutting double eyelids, the current technology can achieve accurate recognition; but if it's because of an accident that leads to a large area of disfigurement, face recognition can't do anything, which requires the fusion application of a variety of biometric recognition technologies (such as live detection, iris recognition, etc.) to further improve the overall efficiency of identity recognition Security.
Face recognition system also has different countermeasures for the behavior of stealing face pictures in an attempt to "muddle through", such as using a video instead of a single photo for recognition, sharing face recognition with other authentication modes, etc. Of course, risk and security are relative, even the more popular fingerprint identification technology, also has the risk of fraud, but because of its convenience is still widely used in daily life.
Whether the biometric technology, including face recognition, can completely replace the traditional methods such as passwords and keys, depends on the actual needs of technology implementation. The progress of technology is not only the improvement of algorithm, but also the actual value that technology can bring us.