Gabor filter responses have successfully used in various important computer vision tasks

such as in texture segmentation

face detection

and iris pattern description. It is evident that Gabor filters have many advantageous or even superior properties for feature extraction. But if the computational complexity cannot be improved their application areas will remain limited. How quickly and accurately using Gabor filter was the identification of the characteristics to become the focus of current research. The paper present Gabor filter envelope based face recognition only the Gaussian part of the filter has to be taken into account; the envelope is similarly the smallest area which includes certain percent of the total filter energy

outside this area can be discarded with only negligible effect in accuracy. The effective envelope is an ellipse which can be encapsulated by a minimal size rectangle. The size of the rectangle may significantly reduce the computational complexity in the spatial domain filtering and save memory in the frequency domain filtering. Experiments using ORL and Yale database indicate that the improved method accuracy outperforms Eigenface and Fisherface algorithm. The new algorithm saves time 20% than traditional methods and gets satisfactory results.