Soft Computing Based Comparative Model for the Classification of Facial Expression Recognition

Classification is a significant step in many applications like image classifications, text recognition, categorization of speech, facial expression classification and so on. And the features fed to the classifier affects the accuracy or performance. The components or patterns of an item in a picture that assist to identify it are termed as features of an item. In computer vision and image processing, the feature carries the information about the content of an image. In classification challenges, the process of extracting features from an object is essential. There are several techniques or features which are used. This paper gives an explanation of different types of feature extraction techniques or methodologies that are used to extract the features out of an image along with classification. Here the facial expressions images are taken. Support vector machine and K-Nearest Neighbor classifier are taken to classify the facial expressions images. The image features used here are local binary pattern (LBP), entropy and histogram of oriented gradients (HOG). The CK+ 48 image dataset is taken here and the MATLAB software is used to obtain the results. The features are extracted and fed to the classifiers. The results show a better accuracy when one feed the three features (LBP, HOG, and entropy) simultaneously. According to the results it is concluded that the accuracy or performance of a classifier depends on the selection of the features of an image or we can say that the classification results are dependent on selected features and classifiers.