In this paper, the system consists of many steps, the first step includes the histogram equalization, detection, feature extraction, and classification. At first, the data set of a face image is segmented into four segments, after that Local Binary Pattern (LBP) algorithm is performed to extract features for each segment. The best feature vectors for all persons are stored in a new dataset in the next stage in order to be used in the testing phase. Finally, the accuracy rate of performance is evaluated to prove its robustness. Experiments show satisfying results and more accuracy achieved by the paper.