Facial expression recognition by using crossing point distribution

Due to a problem of current research occurring when recognizing facial expressions from a 2.5 D partial face data set taken from any viewpoint ranging from -45deg to +45deg, we propose a novel method for recognizing facial expressions from a 2.5D partial face data set. A 2.5D partial data set is captured from any viewpoint between -45deg and +45deg. The proposed method is developed for subject-independent facial expression recognition. To recognize facial expression, a 3D virtual expression face is reconstructed from a 2.5D partial face data set. A facial expression is then represented in terms of the change of crossing points on a face plane. The face plane is divided into 196 (14times14) region partitions according to a crossing point distribution. Numbers of the crossing point in the 196 region partitions are used for recognition by mean of a support vector machine (SVM). The experiments were done for four facial expressions (neutral, anger, surprise and smiling) of 22 persons. The recognition accuracy is 60.9%.