Face Recognition by Feature Extraction and Classification

This article proposed a combination of two algorithms, principal component analysis (PCA) and support vector machine (SVM), which was developed for face recognition. In the proposed technique, the face images were compressed into matrices using PCA and classified using SVM. To test and evaluate the performance of recognizing human faces, this project selected three face image databases, including Yale database, ORL database and Extended Yale Face Database B. All the face recognition rates were reasonable, and the highest face recognition rate was 98.75% for ORL database which subjects to training set number is 8 and 222 dimensions.