A trajectory restoration algorithm for low-sampling-rate floating car data and complex urban road networks. Issue 4 (3rd April 2021)

• Record Type: Journal Article
• Authors: Li, Bozhao; Cai, Zhongliang; Kang, Mengjun; Su, Shiliang; Zhang, Shanshan; Jiang, Lili; Ge, Yong
• Is Part Of: International journal of geographical information science. Volume 35:Issue 4(2021)
• Page Start: 717
• Page End: 740
• Publication Date: 2021-04-03
• Subjects: Trajectory restoration -- map-matching -- A* shortest path algorithm -- low-sampling-rate FCD -- complex road networks; Geography -- Data processing -- Periodicals; Information storage and retrieval systems -- Periodicals; Géomatique -- Périodiques; Systèmes d'information -- Périodiques; 910.285
• DOI: 10.1080/13658816.2020.1825721
• Languages: English
• ISSNs: 1365-8816

Abstract

ABSTRACT: Low-sampling-rate floating car data (FCD) are more challenging than those with high-sampling-rate FCD for map matching (MM) algorithms. Some MM algorithms for low-sampling-rate FCD lack sufficient efficiency nor accuracy, especially related to complex urban road networks. This paper proposes a new method named the trajectory restoration algorithm, which is based on geometry MM algorithms to ensure efficiency and accuracy. The proposed algorithm adopts the modified A* shortest path algorithm to reduce the number of function calls and fully considers road network topology and historical matched points to improve its accuracy. We test the efficiency and accuracy of the trajectory restoration algorithm with FCD data for the complex urban road networks in Beijing. The results have strong continuity which greatly improves the utilization of FCD. We show that the proposed algorithm outperforms related MM methods in efficiency and accuracy and its robustness to restore trajectories of both high and low sampling rates in complex urban road networks.