Design Of Face Recognition System On Cortex-A9

In the wake of the continuous progress of science and technology,we have entered the era of networking and informatization.People are increasingly demanding security and portability in various situations in life.Face recognition as a biometric recognition technology has been widely used and developed in various occasions due to its unique advantages in the context of the rapid development of computer technology and pattern recognition technology^[1].Experts and scholars at home and abroad have poured into the research on face recognition technology.But most of their researches are based on the PC platform.Although face recognition technology based on PC platform has many applications in people's daily lives,they are limited in many occasions because of their large power consumption and large size so that they are not easily portable.Because the many advantages of smart mobile devices,they are widely used in life.The research and application of embedded face recognition systems become more and more important.This article designs a face recognition system,which uses Cortex-A9as the embedded development platform.The platform has many advantages such as low system power consumption,small size and easy portability.The algorithm design of face recognition system mainly involves the realization of face detection algorithm and face recognition algorithm.This system uses face detection algorithm based on Adaboost and face recognition algorithm based on the PCA to build a portable embedded face recognition system on hardware development platform of the Cortex-A9.This system uses the i TOP-4412 development board as a hardware development platform.Compilation and installation of OpenCV library and Qt library and configuration of Qt integrated development environment are completed on the host machine and target machine respectively.Then the modules of image acquisition,image preprocessing,face detection and face recognition included in the face recognition system are theoretically studied and implemented.After realizing the function of face recognition on the embedded platform,the performance indicators such as recognition rate and recognition speed of the system were tested.The results show that the embedded face recognition system based on Cortex-A9 is stable and has a high recognition rate.Therefore,this system will have very good application prospects.