A human face provides a lot of information that allows another person to identify characteristics such as age, gender, etc. Therefore, the challenge is to develop an age group's prediction system using the method of automatic learning. The task of estimating the human age group from images of your frontal face, but it is challenging because of the pattern of personal and non-linear aging which is different from one person to another. Based on presenting face image with accuracy, examines the problem of predicting the age group of humans. The purpose of this study is to prepare a framework and later an algorithm that helps in estimating age group with proper accuracy of face images. In this paper, we present a method for prediction by age group, in which the age group is predicted by detecting face or face reference points using the violo-jones algorithm. After detecting the face, the features include geometric characteristics, wrinkle characteristics and HOG characteristics, and then these extracted features are used to train a classifier using neural networks. The system used self-creation databases for age group classification. In the end, the identification rate obtained by the HOG-neural network model makes better results.