Face recognition algorithm for person identification based on multi-layer information fusion convolutional neural network

A small-scale face recognition network structure is proposed based on Principal Component Analysis (PCA) and Convolutional Neural Network (CNN), to solve the problem that the large-scale network structure has a long running time on low computing power android board. PCA is used to reduce the dimension of features extracted from convolution layer and finally those features after dimension reduction are merged together to make extracted features more compact and more discriminative, besides, the increase in computation comes with information fusion is negligible. And in the input picture fusion its Local Binary Pattern (LBP) image, thereby reducing the influence of light and keep the details of the face as much as possible. The experiments on the Olivetti Research Laboratory (ORL) database of faces show that method proposed in this paper is improved from 80% to 87.5% compared with the traditional neural network. The method proposed in this paper can be used on a low computing power android board.