Face recognition access control system working method

       After nearly 40 years of development, face recognition has made great progress, and a large number of recognition algorithms have emerged. These algorithms cover a wide range of topics, including pattern recognition, image processing, computer vision, artificial intelligence, statistical learning, neural networks, wavelet analysis, subspace theory, and manifold learning. So it is difficult to classify these algorithms with a uniform standard. According to the different forms of input data, it can be divided into face recognition based on still images and face recognition based on video images. Since the face recognition algorithm based on still images is also applicable to face recognition based on video images, only those recognition algorithms that use time information belong to the face recognition algorithm based on video images.

        The composition analysis of the face recognition system: the system consists of the front-end face capture acquisition subsystem, the network transmission subsystem and the back-end analysis management subsystem, which realizes the collection, transmission, processing, analysis and centralized management of the traffic face information. In the system, the front-end face-collecting device is responsible for the collection of face images, and the access server mainly implements the function of receiving and forwarding pictures and information, and can provide unified access services for the capture machines of various models and multiple manufacturers, and receives the received services. The captured image is stored in the cloud storage unit, and the captured image and image are modeled by the face structured analysis server and the blacklist real-time comparison alarm is performed, and the modeled face information and the model data are stored in the big data unit. The back-end parsing application platform supports real-time face capture and retrieval functions according to the user's application needs, and can provide real-time comparison information between the blacklist library and the captured image to provide users with services for quickly and efficiently detecting suspicious targets.

       1. Face detection: Identify the face of the person in the picture, find the position of the face, and circle it

       2. Face alignment: On the basis of the detected face, the iconic feature position such as the nose, mouth and face of the person's face is automatically found.

       3. Face verification: determine whether two faces are the same face

       4. Face recognition: determine who is the face of the face in the picture