Maritime Route Extraction with Proximity Filtering and Clustering for Ship Trajectories

Maritime route extraction through trajectory mining plays a vital role in enhancing safety, efficiency, security, and environmental sustainability within the maritime transportation industry. This research paper explores the use of Automatic Identification System (AIS) data to extract maritime routes. The process begins with a meticulous pre-processing phase, where trajectory data is cleaned to remove outliers and correct errors. This step is crucial for ensuring the accuracy and reliability of the dataset used in subsequent analyses. Following this, proximity filtering is employed to improve computational efficiency. This technique filters trajectories based on their proximity to a specified area of interest, which reduces the dataset size and focuses the analysis on relevant trajectories. To identify and extract maritime routes, various clustering algorithms are tested. Among these, MD-DBSCAN (Modified Density-Based Spatial Clustering of Applications with Noise) and DBSCAN (Density-Based Spatial Clustering of Applications with Noise) are evaluated. MD-DBSCAN, which integrates both spatial distance and Course Over Ground (COG) angle, is found to be the most effective clustering algorithm for this application. The incorporation of COG angle enhances the algorithm’s ability to group trajectories more accurately, leading to better identification of maritime routes. The efficacy of the MD-DBSCAN algorithm is demonstrated through its superior performance in clustering quality and accuracy of route identification. This approach not only optimizes the route extraction process but also significantly reduces computational time compared to other methods. The results of the clustering are assessed using three distinct metrics, and visualizations are provided to illustrate the effectiveness of the clustering approach and the accuracy of the identified maritime routes