Face recognition deploy and control system can facilitate recognizing criminal suspects rapidly and effectively, which is an important means of safeguarding social security. In this paper, a dynamic face recognition deploy and control system based on deep learning framework is put forward, giving a system architecture specific to small and large organizations; the deployment requirements of face camera in the system are given out from the aspects of actual installation and physical lens selection; the requirements of the quality of images in the database and of snapshot images are detailed. Tests show that the system can accurately recognize faces, including the case where the face is partially covered, and the recognition accuracy is higher than 95%. Presently, the system has been put into use, and it works well.
展开▼