期刊名称:Faculty of Computer Science and Information Technology
出版年度:2011
卷号:0
期号:0
语种:English
出版社:Faculty of Computer Science and Information Technology
摘要:A face image consists of a variety of information that makes it recognizable, one facial expression. Some basic facial expressions grouped by the psychological expression of which is neutral and happy / happy / smiling. Although the form / display different expression of each individual, human remains can be recognized. Facial expression recognition system made two expressions intended to recognize the facial expression neutral and happy expressions / smile. After passing the detection area of the face image of face detection results will be extracted using the method of Principal Component Analysis (PCA) which will reduce the dimensions of the face image. At the introduction stage use of artificial neural networks Back-propagation algorithm. Facial image has been reduced will be the input for artificial neural network previously trained with some of the training patterns. Then the artificial neural network to recognize facial expressions in the image by calculating an output. If the value of artificial neural network output close to 0 then the image recognizable as a face image with neutral expression, whereas if the value of artificial neural network output close to 1 then the image of recognizable faces in the image with expression of joy / love / smile.