06-03-2017, 11:03 AM
An experiment for the facial recognition problem by combining eigenfaces and neural networks. The eigen faces are applied to extract the relevant information in an image of the face, which are important for identification. Using this we can represent face images with several coefficients (about twenty) instead of having to use the whole image. Neural networks are used to recognize the face through learning the correct classification of the coefficients calculated by the algorithm of the eigen face. The network is trained first in the images of the database of the face, and then used to identify the images of the face that are given. Eight subjects (people) were used in a database of 80 facial images. A recognition accuracy of 95.6% was achieved with vertically oriented front views of a human face.
The face recognition system consists of two parts: hardware and software. This system is used for users of automatic recognition or confirmation of password. For the input digital images or video frame of the same video are used. The state institution and some private organization are using these systems for facial recognition especially for the face of identification by video cameras as the input parameter or for the biometrics system to verify identity using cameras and 3D explorers . The system must recognize where the face is in an image, remove it from the image and do the verification. There are many ways to verify, but the most popular is the recognition of the characteristics of the face. The face has about 80 characteristic parameters some of them are: width of the nose, space between the eyes, eyehole high, zygotic bone shape and width of the jaw.
The face recognition system consists of two parts: hardware and software. This system is used for users of automatic recognition or confirmation of password. For the input digital images or video frame of the same video are used. The state institution and some private organization are using these systems for facial recognition especially for the face of identification by video cameras as the input parameter or for the biometrics system to verify identity using cameras and 3D explorers . The system must recognize where the face is in an image, remove it from the image and do the verification. There are many ways to verify, but the most popular is the recognition of the characteristics of the face. The face has about 80 characteristic parameters some of them are: width of the nose, space between the eyes, eyehole high, zygotic bone shape and width of the jaw.