الفهرس 
الغلافOnly 14 pages are availabe for public view
 |  | from | 115 |  |  |
| | | from | 115 | | |
Abstract
The topic of automatic recognition of facial expressions deduces a lot of researchers in the late last century and has increased a great interest in the past few years. Several techniques have emerged in order to improve the efficiency of the recognition by addressing problems in face detection and extraction features in recognizing expressions. This thesis designed and implemented an automatic system for facial expression recognition. This system consists of face detection phase, facial feature extraction phase and recognition phase. In face detection phase the system has utilized Viola & Jones technique to detect of face. In feature extraction phase the system has used hybrid approach which is represented as a combination between holistic and analytic approaches by the extract of 307 facial expression features (19 features by geometric, 288 feature by appearance). Expressions recognition is performed by using radial basis function (REF) based on artificial neural network to recognize the six basic emotions (anger, fear, disgust, happiness, surprise, sadness) in addition to the natural. The system has achieved a recognition rate of 97.08% when it is applied on person-dependent database and 93.98% when applied on person-independent database.