Intelligent Face Recognition System Based On CORTEX-A8Embedded Technology

With the development of Internet of Things and Smart Home, Biometric identificationtechnology receives more and more attention. Among all the kinds in Biometric identificationtechnology, face recognition stands out because of its stability and individual differences,which makes it a popular field of authentication.First, we analyze the status of the study for face recognition in domestic and foreign andthe problem of recognition rate, according to which, we resign the task of this paper. Then weintroduced the related theory and current algorithms of face detection and face recognition.We have a detailed description of Concepts and definitions the of basis of mathematicalmodels-Hidden Markov(HMM) and based on Embedded Hidden Markov(EHMM), listing thethree basic problems of EHMM.We use Adaboost algorithm to implement face detection and face recognition based onEHMM. The four steps are establish model, feature extraction, train and face recognition. Inthe module of feature extraction, we improve the traditional algorithm--Dimensional discretecosine transform(2D-DCT) to Local Binary Pattern(LBP) and point out its advantage. Wedescribe the design and implementation of the whole scheme. According to many testing, wefind out the features of our algorithm and contrast LBP with2D-DCT.Finally, we conclude the advantage and disadvantage of our system, look ahead thedevelopment of face recognition in the future and the problems needed to solve.