 The proposed method improves the accuracy of merging or updating road networks by considering both local and global optimizations, inserting virtual nodes to identify MN matching patterns, and using attribute-related evaluation indicators to evaluate matching accuracy. Experiments show that the overall matching accuracy exceeds 90 percent, increases by 6 to 12 percent after an MN pattern is added, and by 4 to 6 percent following the addition of topology indicators. The proper buffer threshold is about twice the average value of the clasest distance from all nodes. This article was authored by Jianchen Zhong, Yan Hui Wang, and Wenji Zhao.