Human facial Expression recognition through computer by means of high recognition rate and less time consumption is still a demanding and motivating task. Various emotional information is conveyed by facial expression alone. Facial expression recognition system have a vital role in numerous areas such as human-computer intelligent interaction system. This paper aimed to present an Automatic Facial Expressions Recognition System that would recognize five foremost expressions, named as Happy, Sad, Neutral, Angry and Disgusted. In this system recognition of five different expressions is done by using some extracted features. A well known Viola Jones face detection method is used for the detection of the frontal face. After face detection ROI (region of interest) that is eyes and mouth is taken for feature extraction in which local binary pattern (LBP) is used as a feature. Then obtained LBP features are clustered by using an efficient method named adaptive Neuro Fuzzy Classifier (ANFIS) to efficiently recognize various expressions. The whole system is implemented on the dataset of 110 images of frontal facial expressions of Happy, Sad, Neutral, Angry and Disgusted from live Indian facial expression images as well as 54 untrained images have been used for testing the developed system to determine the recognition efficiency of new or untrained images by using MATLAB R2012(b). The database is created from live Indian images. After the successful testing with the proposed system the expression recognition efficiency found is 85.45% for five specified expressions for trained images and 38.89% for untrained images.