preprocessing

Shipping route analysis for maritime traffic management depends on equipments to collect information about ship's behavior. For this purpose, the most reliable data is the Automatic Identification System (AIS) data. The complexity and high volume of AIS data  enhances traditional surveillance operations and makes maritime traffic analysis more difficult. Therefore, an unsupervised approach is desirable for the effective conversion of raw AIS data into regular shipping route patterns. The proposed model for the shipping route analysis consists of four sections: AIS data preprocessing, structural similarity measurement, shipping route clustering and representative trajectory extraction. Experimental evaluation of the proposed model with real AIS data from the studied area shows that it has performed well visually and the expected result has been achieved. The results will contribute to better understanding of shipping route patterns and help maritime authorities in sustainable management of maritime traffic.