The invention provides a processing feature recognition and information extraction method for an MBD model. The method comprises the following steps: 1, analyzing the MBD model; 2, traversing the cylindrical surface, and identifying hole characteristics; 3, initializing an attribute adjacency graph storage matrix, and constructing an attribute adjacency graph; 4, traversing predefined processing features and feature sub-graphs thereof; 5, identifying all the machining feature surface sets of the type by using a sub-graph isomorphism method; 6, calculating the key geometric dimension accordingto the processing feature type; 7, extracting three-dimensional labeling information of the model; 8, extracting part annotation information; 9, interactively identifying self-defined processing characteristics; 10, visualizing a machining feature recognition result; 11, outputting a feature recognition and information extraction result in a structured manner. Through the steps, the machining feature surface set in the MBD model can be recognized, the key geometric dimension of the machining feature surface set can be automatically calculated, the three-dimensional annotation and annotation information of the part can be extracted, the final result is output in an XML form, and a good data foundation is laid for subsequent process design.