The Convolution Neural Network (CNN) is a type of deep-run learning that is composed of neural networks. It is especially popular in the field of image recognition and speech recognition. It is also possible to process data in a multidimensional array, such as an image. In this study, we propose a system based on artificial neural network which predicts through the learning of sectional image of aircraft airfoil. The NACA series is used for the prediction for the airfoil, and the shape changes according to the size of the leading edge and trailing edge. In order to check the shape of the airfoil, it is necessary to use the coordinate value of the shape or a specialized program. In this study, the feature factors were derived from the learning image using the convolution neural network. The machine learning was conducted through the derived factors and the sectional shape of the aircraft airfoil could be predicted only by the specification of the airfoil.