Research on Face Recognition Algorithm in Embedded System

As a friendly biometric method, face recognition has the advantages of not easy to forge, easy to obtain, and high accuracy. The paper builds a development environment for embedded face recognition. In terms of hardware, the embedded development board and camera of the ARM architecture are chosen; in terms of software, the open source Linux operating system is chosen. The virtual machine was installed on the PC, the cross-compilation environment was established, and the kernel, driver and other related ports were transplanted on the development board, which established a stable running environment for the design and development of the face recognition application. An embedded system pooling-patch algorithm based on deep learning pooling operation and Deep ID algorithm is proposed. Experiments on the ORL face database and the face image library of our lab members show that the method not only has a high recognition rate but also takes a short time.