Abstract: Objective Corresponding feature matching,essentially as a matter of global combinatorial optimiza-tion,is one of the key technologies for geospatial data integration,fusion and update.In this paper,aglobal optimum matching solution is achieved taking the advantages of ant colony optimization groupsand random search,without the centralized control and global model.The basic principle of ant colonyoptimization for road network matching is explained first,with a mathematical constraint model con-sidering both geometric error and structural characteristics.Then,the matching problem solutionmodel is designed,with a self-adaptation and local search strategy employed to improve efficiency.Fi-nally,the key steps are given.Experiments show that the ant colony optimization approach is effec-tive,feasible and practical,providing a new idea for road network matching.