Small sample classification learning method based on graph neural network

Contact network defect detection based on a 4C image plays an important role in ensuring railway transportation safety and stable operation of a train, and a method based on deep learning has the problems that the number of parts and defect types is large, and the number of defect samples is small. The invention provides a small sample classification learning method based on a graph neural network, and the method comprises the steps: 1) collecting a real scene high-speed rail overhead line system picture, carrying out the positioning and classification processing of parts, and constructing a 4C sample library; 2) inputting the pictures of the corresponding categories into a small sample classification network for training to obtain a model training result; 3) obtaining a corresponding category name of a single unmarked test picture through the trained model; and 4) comparing predicted values with true values of the marked test pictures input in batches to obtain specific classification precision. According to