A NOVEL FUSION BASED HYBRID APPROACH FOR FACE RECOGNITION SYSTEM

摘要

Face recognition plays a challenging role in the field of image processing. Though one of its benefit is that the person can easily be identified among the crowd through face recognition, but it has several issues that need to be overcome such as facial expressions, ageing, occlusion etc. Face recognition method can only be entertained, if it can identify the face correctly. Global features (DWT, PCA) and point based features (Harris corner, SURF) both are considered to create a fusion based system for automatic face recognition. Accuracy, sensitivity and precision are the performance evaluation parameters are evaluated. The comparative study of various standard and conventional feature extraction techniques is carried out in this paper in order to decide which approach is more suitable for accurate face recognition. Fusion based approach gives 99.4 percent accuracy.