An Improved Map-Matching Method Based on Hidden Markov Model

Map-matching is the process to match a sequence of real world coordinates into a digital map, so as to identify the correct segment on which a vehicle is traveling and to determine the vehicle location on the segment. Map matching is one of the key components to model and analyze floating car data, and provide ITS services such as traffic condition analysis and navigation. Complex environment, inadequate attribute information, low sampling frequency, and location deviation exert great influence on the matching performance. This paper presents an improved map-matching algorithm based on Hidden-Markov model. A distance based weighted-average method is applied to improve the quality of the instantaneous GPS data, and a preprocessing and caching method for the shortest paths is used to accelerate the calculation of state transition probability. Comparative analyses show that more than 90% of positions are matched, and computation time is significantly improved.