Detection method and system for resisting attack

The invention provides a detection method and system for resisting attacks, and the method comprises the steps: building a graph model of at least one dimension based on an original image after obtaining the original image of a user, inputting the graph model of at least one dimension into a graph neural network model, determining the related region features of biological features, and carrying out the recognition of the related region features. And inputting the associated region features into an attack-resisting detection model to obtain an attack-resisting detection result of the user. According to the method and the system, the graph model of at least one dimension is utilized to establish the incidence relation between different regional features, so that attack-resistant detection can be carried out from the dimension of the original regional features, and attack-resistant detection can also be carried out from the dimension of the incidence relation between different original regional features; therefo