Neural Network Based
Face Recognition using Principal Component Analysis
Abstract:
Neural networks are used in almost all the fields. From pattern recognition to military purposes neural network has its own importance. Face recognition is an emerging field today. A network is trained to recognise an individual’s face. Now when a set of images are presented to the trained network it will recognise the particular individual’s face. This finds its application in the forensic department.
The faces used for the recognition process are vertically oriented frontal view with wide expression change. They are extracted from the images by the face detection code first. This will eliminates background influence as much as possible. The face space is described by a set of eigenfaces. Each face is efficiently represented by its projection onto the space expanded by the eigenfaces. A new descriptor is then assigned to each face. Neural networks are used to recognize the face through learning correct classification of these new descriptors. A real-time system can be created which combines the face detection and recognition techniques. The images are taken using a camera. The face in the image is then detected and it is recognized. This face database can be easily expanded to accommodate more individuals.
No comments:
Post a Comment