Rapid detection of defect structures in graphene by the machine learning

Graphene has long been considered to have a wide application perspective in many fields, such as electronic information, chemical industry and aerospace. However, the actual graphene materials contain a variety of defects that affect their properties. Accurate and rapid identification of defect structure types is of great significance for the application of graphene. First, the image preprocessing technology is used to improve the classification precision. Next, use the convolutional neural network to classify the type of vacancy defect and perform rapid analytical tasks to accurately identify graphene defect structures. This study aims to contribute to this growing area of research in graphene to classify and count the vacancy defects automatically and quickly in monitoring its surface structures.