A New Optimized Approach to Face Recognition Using EigenFaces
Sheifali Gupta, O.P. Sahoo, Ajay Goel, Rupesh Gupta · Global Journal of Computer Science and Technology · 2010
Eigenface approach is one of the simplest and most efficient methods for face recognition. In eigenface approach chosing the threshold, value is a very important factor for performance of face recognition. In addition, the dimensional reduction of face space depends upon number of eigenfaces taken. In this paper, an optimized solution for face recognition is given by taking the optimized value of threshold value and number of eigenfaces. The experimental results show that if the threshold value is 0.8 times of maximum value of minimum Euclidian distances of each image from other images, then maximum recognition rate is achieved. Also only 15% of Eigenfaces with the largest eigen values are sufficient for the recognition of a person. Best optimized solution for face recognition is provided when both the factors are combined i.e. 15% of eigenfaces with largest eigen values are selected and threshold value is chosen 0.8 times maximum of minimum Euclidean distances of each image from all other images, it will greatly improve the recognition performance of a human face up to 97%.