A Trajectory Ensemble-Compression Algorithm Based on Finite Element Method

• Main Authors: Haibo Chen, Xin Chen
• Format: Article
• Language: English
• Published: MDPI AG 2021-05-01
• Series: ISPRS International Journal of Geo-Information
• Subjects: trajectory compression; ensemble compression; finite element method
• Online Access: https://www.mdpi.com/2220-9964/10/5/334

Trajectory compression is an efficient way of removing noise and preserving key features in location-based applications. This paper focuses on the dynamic compression of trajectory in memory, where the compression accuracy of trajectory changes dynamically with the different application scenarios. Existing methods can achieve this by adjusting the compression parameters. However, the relationship between the parameters and compression accuracy of most of these algorithms is considerably complex and varies with different trajectories, which makes it difficult to provide reasonable accuracy. We propose a novel trajectory compression algorithm that is based on the finite element method, in which the trajectory is taken as an elastomer to compress as a whole by elasticity theory, and trajectory compression can be thought of as deformation under stress. The compression accuracy can be determined by the stress size that is applied to the elastomer. When compared with the existing methods, the experimental results show that our method can provide more stable, data-independent compression accuracy under the given stress parameters, and with reasonable performance.