Pattern of face is recognized by using back propagation algorithm.

Abstract

Face recognition is an effective way for the personal identification of human being. Person identification is based upon reliable, unique, stable or behavioral characteristics. In this paper face print is used as new method for the identification and verification of human being. Functionally Face print recognition is dived into face detection, face matching and face recognition. Implementation of face detection systems is based upon deformable template algorithm which is based upon image variant. Implementation of deformable template algorithm is based upon preceptor. To create the Perceptron's A-units, Kohonen Feature Maps of unsupervised learning is used. Efficiency of face detection model is improved by natural symmetry of faces. The deformable template was run down the line of symmetry of the face in search of the exact face location. Principal Component Analysis is used to realize automated face recognition, which is called as Karhunen-Loeve transform. The implementation of automated face recognition system is tested by using Manual face detection. Under controlled conditions the also successfully implemented recognition is Pose invariant face recognition. Keywords: Karhunen-Loeve transform, Perceptron, unsupervised learning