Decoupling view-related and view-unrelated features is crucial for effective person re-identification in aerial-ground scenarios.