LBP known as The Local Binary Pattern non-parametric descriptor which is used to investigate different local structures in an image. It is rendered as a straightforward and effective texture operator for different real-time applications. LBP generated binary data is sensitive to noise and it will identify face images when its completely Visible. To increase LBP's efficiency and resilience, in this research paper we proposed a novel multiple mean fuzzy LBP(MMFLBP)to classify facial images with different facial expressions and rotation invariant images. GLCM with six features is used for construction of Database on the outcome of MMFLBP. Confusion matrix is used for estimating the accuracy of MMFLBP. Upon calculation of the experimental results, it is obvious that our proposed method exhibits better accuracy.