Face Recognition Method Based On Specific

The person face recognition that is an active research area, is one ofhuman vision most outstanding abilities. Although the person facedistinguished rate of accuracy is generally lower than the iris, fingerprintrecognition, but as a result of its non-violation and most nature anddirect-viewing way to the user, the person face recognition is the mostlyaccepted biological features recognition way. In this paper, we propose anovel face model to explicitly characterize the image differencesconveyed by each face image. It exploit knowledge of which types of facevariations are critical for recognition, which are more important inreal-world applications. In the preprocessing stage of face recognitionsystem using this method, we can effectively improve the recognition ratein the equal time.Two face image databases, namely ORL, and Yale, are used toevaluate performance. Experimental results demonstrate the efficacy ofthe proposed specific area selection approach in face recognition.