Human face is an integral part for delivering an amount of nonverbal information to facilitate communication. One of the important is human age, but accurately extracting human age from his face is not easy. Facial aging process is different on early growth (child to adult) and adult aging (adult to senior). On early growth and development of the face, from birth to adulthood, the greatest change is the shape change. But on adult aging the most significant change is on the skin (textural change), the shape change still continue but much less dramatic compared to early growth. In our research we develop a two stage facial age classification system to classify human age into age classes. This paper is the first part of our two stage facial age classification. In this paper we develop facial age classification system for the early growth stage, we classify human age into six classes which are 0-2, 3-4, 5-6, 7-10, 11-16, and 17+. Because the early growth have grater impact on shape change, we only use geometrical feature to classify the human face. We use Haar-like cascade to perform the face detection and then we use Active Shape Model to extract the geometrical features on human face. For the classifier we use Support Vector Machine(SVM) classifier with Radial Basis Function (RBF) kernel. This system is developed with OpenCV library and C++ language. The results of classification based on geometrical features achieved an accuracy rate of 71.25% using FG-NET dataset.