Name
Abstract | ||
|---|---|---|
In recent years, the traffic volume on the sea has increased significantly. Compared with road traffic management, sea traffic management is very difficult due to many reasons. For safety sailing, automatic identification system (AIS) has been introduced. Using AIS signals, it is possible to understand the position, velocity, and other information of each sea-going ship, and thus can detect possible dangers and provide necessary rescue promptly. However, some ship owners may not set their AIS correctly, and thus the AIS signals may not be trustable. The purpose of this study is to propose a method to classify the true ship type using the AIS signal and provide a way to support traffic management. Specifically, we extract the "signature characteristics" of the ship from its AIS signal, and then classify the ship type using a machine learning model. Primary experimental results show that the average accuracy is about 87.3% if we use a multilayer perceptron. Better results are expected if we use more data. |
Year | DOI | Venue |
|---|---|---|
2019 | 10.1109/ICAwST.2019.8923540 | 2019 IEEE 10th International Conference on Awareness Science and Technology (iCAST) |
Keywords | DocType | ISSN |
Behavioral analysis,ship classification,sea traffic management,AIS,machine learning | Conference | 2325-5986 |
ISBN | Citations | PageRank |
978-1-7281-3822-0 | 0 | 0.34 |
References | Authors | |
0 | 2 | |
Name | Order | Citations | PageRank |
|---|---|---|---|
| Shoichi Ichimura | 1 | 0 | 0.34 |
| Qiangfu Zhao | 2 | 214 | 62.36 |