Face Recognition System for Increased Feature Extraction

Efficient feature selection for improved face recognition and biometric authentication

Abstract

Feature extraction is the most vital stage in pattern recognition and data mining. In this stage, the meaningful feature subset is extracted from original data by applying certain rules. For reliable recognition, it is desirable to extract appropriate features space, since all the extracted features may not contribute to the classification positively. In this paper, some feature extraction methods and algorithms were studied, compared and means of improving feature selection through dimension reduction was explained. It was concluded that few number of features are usually required and selected for an optimized compression. So that huge amount of data can be reduced to a relatively small set which is computationally faster. Hence, efficient selection of features is a key step of achieving efficient face recognition and biometric authentication.