Abstract— The purpose of this study is building a face recognition system. Wireless technology and Adaboost algorithm is applied in this embedded system.The whole system includes two parts: server part and client part.The server part, which is based on ARM9 microprocessor, can detect faces, compress the image data and send it to the server part using Wi-Fi wireless transmission module. Then the client part receives the data, manages to recognize face automatically and identify faces by comparing image in the database. The client part deals with face detection and face region positioning via Adaboost fast face detection algorithm combined with digital image processing of grayscale balance, filtering, wavelet transform to acquire standard face image. Three-layer BP neural network is designed for automatic face recognition. The result achieve a recognition rate at 98.6 percent with a small number of 45 training samples.

Index Terms— Linux, Wi-Fi, Adaboost, BPNN.

Mi Wang, Kezhong Zhang, and Peizhi Liu are with the Being University of Posts and Telecommunications, Beijing, China (e-mail: ｛wangmi666, zkz.bupt｝@gmail.com, liupz@bupt.edu.cn).

Cite: Mi Wang, Kezhong Zhang, and Peizhi Liu, " Research and Design of Face Recognition System Based on Embedded Wi-Fi," International Journal of e-Education, e-Business, e-Management and e-Learning vol. 3, no. 6, pp. 497-501, 2013.