Facial Expression Recognition has been a very essential area for research in pattern acknowledgement and presently, there is no method of a facial expression recognition structure that has a 100% recognition rate or say a 100% efficiency. So find the problems are to increase recognition rate by improving the preprocessing of datasets, and also refining the feature extraction method with using the best classifier for facial expression recognition. Feature extraction is the significant step on which recognition rate totally depends on facial gesture recognition. High dimension and high redundancy are a problem for Gabor while it has an extreme variance of features. Dimension and redundancy should be compact using the filtering practice. In the proposed Gabor feature extraction method, the Gabor features are clarified using wavelet change and obtained optimum features on the facial Gabor matrices.

KEYWORDS: DWT, Facial Expression Recognition, Gabor Filter, Gesture.