Face recognition based on neural network ensemble and feature fusion

In this paper, a fusion of multiple features method based on integrated neural network was proposed. Firstly, KPCA algorithm was used to extract the overall recognition features. Then, the KICA algorithm was used to extract the local features. Finally, the BP and PNN neural network were used to recognize the result of the both algorithms. The method solved the interference caused by the external factors and got high correct rate of classification.