Abstract

This paper presents a hybrid framework for face recognition based on the component features that are extracted by the salient points from the facial components of the face. Face recognition is still challenging under different pose, illumination conditions etc. The Proposed method detects the face from the original image. The components (eyes, nose and mouth) are extracted from the detected face through Viola-Jones algorithm. Extract the component features using Speeded up Robust Features (SURF). Extracted features are concatenated into a single feature vector to represent the face and compute the similarity between the query face and feature vectors stored in the database. The proposed hybrid method is implemented on FERET and ORL datasets and it improves the recognition rate and decreases the system running time.