Features Fusion Based on FLD for Face Recognition

Abstract

In this paper, we introduced a features fusion method for face recognition based on Fisher’s Linear Discriminant (FLD). The method extract features by employed Two-Dimensional principal component analysis (2DPCA) and Gabor wavelets, and then fuse their features which are extracted with FLD respectively. As a holistic feature extraction method, 2DPCA performs dimensional reduction to the input dataset while retaining characteristics of the dataset that contribute most to its variance by eliminating the later principal components. On the contrary, the Gabor transformed face images exhibit strong characteristics of spatial locality, scale and orientation selectivity, which produce salient local features which are most suitable for face recognition. So, we use Gabor wavelets for the local features and then integrate Gabor features with 2DPCA features. In addition to, because FLD could make not only the scatter between classes as large as possible, but the scatter within class as small as possible, the features which are extracted by FLD are reliable for classification. And then, the FLD features of the Gabor and 2DPCA features leads to the application of the support vector machine (SVM) for classification. Finally, the computer simulation illustrates the effectivity of this method on three independent face databases.