الفهرس 
الغلاف الانجليزىOnly 14 pages are availabe for public view
 |  | from | 131 |  |  |
| | | from | 131 | | |
Abstract
Since the early 1990’s, face Recognition Technology (FRT) became an active research area. Most of the current profile recognition algorithms depend on the correct detection of fiducial points and the determination of relationships among these fiducial points. Unfortunately, some features - such as concave nose, lips, flat chin, etc. - make detection of such points difficult and unreliable. Also the number and position of fiducial points vary when pose changes even for the same person.
In this thesis, face profile recognition is based on Fusion of various techniques to recognize the human face throughout the side view and to qualify the in use proposed techniques and going into various problems to overcome. The four proposed techniques of the face profile recognition in this work are presented which are as follows: the face profile recognition using Fast Fourier Transform (FFT), the face profile recognition using the curve fitting technique, the face profile recognition using the coordinates, and the face profile recognition using segmentation. The last technique is divided into three segments (A), (B), and (C), which does not require the extraction of all the fiducial points only, but it also uses information contained in the profile.
Each technique contains several stages, the first stage, which is called preprocessing stage one, is used with all proposed techniques. The second and third stages are completely different in the different techniques. After the features are extracted the curvature coefficient values from the curve fitting equations are used and the fiducial points - such as nasion, chin, and forehead - can be reliably extracted using a fast and simple method. The last stage is also common for all techniques used in this thesis and it has two methods for matching and classification.
The first method is the Nearest Neighbor using the Euclidean distance, which is used for each technique. The second method is the Neural Network which is used for two techniques (Fast Fourier Transform technique and Segmentation technique). These two methods are applied to match the face profile values where the tested images are compared to the database images.
Experiments are performed on 50 clients chosen with different ages, in public area. Each had three shots in different time capture and illuminations, getting total of 150 images database. Three techniques are used for comparison as follows: the face profile recognition using the Fast Fourier Transform technique, the Tangency- Based technique, and the Curvature- based technique. Recognition ratios and results are presented and discussed.