Facial Expression Recognition is essential for designing ny human-machine interface. The main concern of Facial Expression Recognition is to come to a decision what features are required to represent a Facial Expression. A research has been carried out in facial expression recognition by solving the research issues of facial expression recognition under different illuminations, orientations and many other variations but no technique so far has achieved a 100% accurate recognition rate. Feature extraction is the main research issue for facial expression recognition and the objective of this paper work is to extract a comparative performance analysis of different feature extraction
technique for facial expression recognition. Experimental are performed on seven expressions, (anger, disgust, fear, happiness, sadness, surprise, neutral) of JAFFE dataset. The features of facial expression image have been extracted by using Discrete Cosine Transform, Gabor Filter, Discrete Wavelet Transform and Gaussian distribution of a facial image and feature vector is then proceed to Adaboost classifier for automatic recognition.
Feature Extraction, Facial Expression Recognition, DCT, Gabor Filter, Wavelet Transform, Gaussian distribution