Face recognition based on Gabor features and kernel discriminant analysis method
Jiliu Zhou · Jisuanji gongcheng · 2005
A Gabor features based classification using kernel discriminant analysis for face recognition was introduced.First an augmented Gabor feature vector was derived from the Gabor filters.Then the kernel discriminant analysis(KDA) method was used for enhanced face recognition performance.The Gabor features characterized by spatial frequency,spatial locality,and orientation selectivity can do with the variations due to changes in illumination and facial expression.The KDA method can derive the nonlinear discriminant features in the input space.The feasibility of this method has been tested by experiments.