Facial Expression Recognition in Various Illuminous Environment

The project presents human emotion recognition from face images based on textural analysis and linear classifier. Usual Facial Expression Recognition plays an important role in Human Communication Interaction systems for measuring people’s emotions has dominated psychology by linking expressions to a group of basic emotions (i.e., anger, disgust, fear, happiness, sadness). The recognition system involves face detection, features extraction and selection and finally classification. The face detection module will be used to get face images which have normalized intensity, are identical in size and shape and illustrate only the face section. The Gabor filter bank is then used to mine the features from face regions to discriminate the illumination changes. The feature selection module is required to select unique features. In previous term, the feature selection section helps develop the performance of learning models by removing most irrelevant and redundant features from the feature space. The finest features are particular using less redundancy more relevance algorithm based on mutual information. The mutual information quotient (MIQ) method for feature selection is adopted to select the optimum features. These features are valuable to differentiate the maximum number of samples accurately and the linear classifier based on discriminate analysis is used to the six different expressions. The simulated results will be shown that the Gabor filter based feature extraction with used classifier gives much better accuracy with lesser algorithmic complexity than other facial expression recognition approaches.