A Practical HMM-Based Map-Matching Method for Pedestrian Navigation
- Resource Type
- Conference
- Authors
- Ma, Shengjie; Lee, Hyukjoon
- Source
- 2023 International Conference on Information Networking (ICOIN) Information Networking (ICOIN), 2023 International Conference on. :806-811 Jan, 2023
- Subject
- Communication, Networking and Broadcast Technologies
Computing and Processing
Navigation
Current measurement
Roads
Urban areas
Hidden Markov models
Markov processes
Trajectory
pedestrian navigation
map-matching
hidden Markov model (HMM)
map-matching initialization
- Language
Map-matching is an important component in pedestrian navigation. The proliferation of pedestrian navigation applications is being limited by the severe inaccuracies of the current GPS available on smartphones, especially in dense urban areas where the distance between neighboring streets and alleys is smaller than the typical GPS error range. Although map-matching methods based on HMM are considered as a practical approach for GPS error correction, there exist a few issues to be addressed before a wide-scale deployment can be made. In this paper, we propose an algorithm to determine the initial probabilities of hidden states using a small number of GPS measurements. An arrival probability is computed for each past GPS measurement, which indicates the probability that a past GPS measurement will arrive at a road segment within the initialization duration. The experimental results show that the proposed map-matching initialization algorithm can effectively determine the initial road segment compared with the traditional HMM-based map-matching methods and increase the accuracy of pedestrian map-matching.