Spatio-Temporal Vessel Trajectory Smoothing Using Empirical Mode Decomposition and Wavelet Transform

The Automatic Identification System (AIS) has attracted increasing attention in recent years for its superior properties in ocean engineering and maritime management. The spatio-temporal vessel trajectory data is highly related to the received AIS data. However, the AIS raw data often suffer from undesirable noise during signal acquisition and analog-to-digital conversion. To improve AIS-based vessel trajectory data quality, we propose to develop a vessel trajectory smoothing method by combining empirical mode decomposition (EMD) with wavelet transform. In particular, EMD is introduced to decompose the original vessel trajectory data into several sub-trajectories. The EMD decomposition is able to assist in enhancing the robustness of trajectory smoothing. Wavelet transform is directly adopted to smooth the decomposed sub-trajectories. The final high-quality trajectory is obtained by combining the smoothed sub-trajectories in this work. The proposed method has the capacity of removing the unwanted noise while preserving the important trajectory features. Numerous experiments have illustrated the superior smoothing performance of the proposed combined method.