I. INTRODUCTION Turing found that uneven patterns or spatial inhomogeneity may come from diffusion, called Turing instability [1]. From a biological point of view, the mottling pattern presented by organisms is determined by genes and chemical reactions in the body, and the classification of mottling is conducive to exploring the theoretical mechanisms and biological roots of mottling formation, which can deepen the understanding of the interactions between planktonic organisms, and even uncover the causes of certain diseases and abnormal expression of genes as a means to prevent diseases. The faster development in the field of image classification is due to the discovery of deep learning. Deep learning algorithms have a greater advantage over traditional machine learning algorithms for image classification problems. In particular, convolutional neural networks mimic the way neurons in our brain recognize patterns in image classification problems.