Non verbal communication has its own benefits and importance from the perspective of manyapplications. Facial expressions are one of the significant forms of non verbal communication amongst othermajor non verbal communication indicators like brain signals, body posture, gesture and actions etc. usedprominently to convey the emotional state/ mood of a person. Emotion recognition serve wide range ofapplications like healthcare, patient pain monitoring, driver alert system, cognitive assessment, e- learning,animation etc. Emotion classifier is modeled around the features fed to the model in the form of feature vectori.e. set of prominent features/attributes. The correctness in extracting the facial features has tremendous impacton classifier accuracy. The paper presents the study of various popular and unique techniques used so far forfacial feature extraction and emotion classification. Various techniques of facial expressions analysis arecompared over the performance parameters like recognition accuracy, number of emotions addressed, Databaseused for experimentation, classifier used etc