A method for detecting a two-dimensional material is provided. In the method, a suitable OM image from a cloud database is converted into a space spectral data by a hyper spectral imaging technology of visible light and a Raman data of a grown layer number and a grown position of a sample are used as a label of data set, and the space spectral data and the label are used as an input of a model with 3D-CNN.A method for detecting a two-dimensional material is provided. In the method, a suitable OM image from a cloud database is converted into a space spectral data by a hyper spectral imaging technology of visible light and a Raman data of a grown layer number and a grown position of a sample are used as a label of data set, and the space spectral data and the label are used as an input of a model with 3D-CNN.With the superior performance of two-dimensional (2D) materials and its wafer-level synthesis methods, it has attracted more and more attention, aroused great interest and triggered a revolution in the application of corresponding devices. However, the large-scale characterization, accuracy, intelligent automation and high-efficiency detection of nanostructures for two-dimensional materials have not yet been successfully applied at industrial level. It can be applied to semiconductor industry, display, photoelectric material industry.