Image classification is an essential and widely used area where deep learning is applied. The deep learning approach has been extensively applied in the area of image classification and provides very good classification accuracy. Some of the deep learning approaches can classify images better than a human. Image classification has tremendous applications in practice. This paper presents a survey on some of the deep learning approach-based image classification methods which have been extensively used in various applications of image classifications. The deep learning approaches which have been considered in our study and are used for developing a variety of image classification methods are Convolutional Neural Networks (CNN), Recurrent Neural Networks (RNN), Long Short Term Memory (LSTM), Generative Adversarial Networks (GAN), Restricted Boltzmann Machine (RBM) and Deep Belief Network (DBN). The paper also discusses comparative studies on some of the image classification techniques which have been used in different areas of image classification problems.