Design of face recognition system based on CNN

Abstract

This paper studies and implements a face recognition system based on convolutional neural network. The system firstly uses AdaBoost algorithm to detect the face, that is to get the position and size of the face accurately in the image, then uses the deep convolution neural network to extract the face features and classify them. Finally, the hardware application Altera DE1_SOC development board, designs and completes a face recognition system with high recognition rate.