A Map Matching Algorithm Based on Road Reduction Filter（RRF） is proposed in this paper

which pays particular attention to the matching problems that arise at intersections. Application of credibilist multi-criteria association using belief theory and Dempster-Shafer’s rule in only two roads map matching problem is studied. Two criteria

the shortest Euclidean distance and the smallest angle difference between road segment and vehicle heading are presented. Moreover

the correlation theory is introduced to simplified MM problems from multiread segments to only double road segments. As a result

the probability assignment function in criteria can be adjusted adaptively in accordance with actual road network. Simulation results with actual field experimental data show that RRF can perfectly solve the MM problem at urban intersections.