A Development of Emotion Recognition System

Human Computer Interaction (HCI) is one of most interesting topic in machine visualization and image processing fields. Emotion recognition plays an important role in security and interpersonal communication. Biometric system helps in identification, security and authentication using face image. Recognize emotion of a person from occluded face image is a challenging task in emotion recognition. Feature are calculated for face image using Principe Component Analysis (PCA) and Two-Directional Two Dimension Principal Component Analysis [(2D) 2PCA] along with discrete wavelet transform. K-Nearest Neighbor (K-NN) and multiclass support vector machine used for classification of different emotion. This paper shows the comparative study of feature extraction and classification method. This study is performed in three dataset. JAFFE, CMU and CK database is used for calculating the classification rate of emotion recognition system .Resulting successful classification rate for JAFFE database is 91.8919% for CMU dataset classification rate is 70.339 % and for CK database resulting classification rate is 75.3073% using Multi class support vector machine. Multiclass support vector machine gives better result as compare to K-nearest neighbor.