A Feature Extraction Algorithm Using Data Fusion in Human Face Recognition System

This paper investigates the feature extraction algorithm in human face recognition systems, especially the principle component analysis and linear discriminate analysis. Based on the data fusion theory, a new method of feature extraction is presented; it utilizes the advantages of both methods. Then the experimental system is constructed that the human face in data space can be projected to human face in feather space using the new feature extraction. The constitution and characteristics of the system is analyzed and the experimental results are given based on the back propagation neural network for face recognition based on the new feature.