Analysis of Facial Detection and Classification Using Various Approaches

This paper proposes a system which automatically recognizes the emotion represented on a face. Initially the face image has been captured and the image is analyzed according to the skin color. After the color image is transmitted to the grey scale image and the noise present image is eliminated with the help of the non-local median filtering approach. From the preprocessed image different features are extracted by using the progression invariant sub space learning method. Then the optimized features are selected and trained for improving the classification rate which is done by using the back propagation neural networks. Finally, the face related features are classified using Support Vector Machine. There are different seven facial expressions considered over here: happy, angry, surprise, disgust, fear, sad and neutral. The system was tested on the Japanese Female Facial Expression (JAFFE) dataset.