Multiple Ornstein–Uhlenbeck Processes for Maritime Traffic Graph Representation

Coscia, Pasquale; Braca, Paolo; Millefiori, Leonardo M.; Palmieri, F.; Willett, Peter

doi:10.1109/taes.2018.2808098

Cited by 62 publications

(27 citation statements)

References 18 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Since each DB-Scan algorithm is performed on each tile, this results in running several thousands of DB-Scan algorithms in parallel (see Section 4). In our case, performance is measured as in (Pallotta et al 2013) and (Coscia et al 2018) as the percentage of AIS positions of the test set that fall within the convex hulls and use the term 'accuracy' to refer to it. Specifically: Table 3 shows the averaged accuracy achieved by our methodology for each vessel type.…”

Section: Resultsmentioning

confidence: 99%

“…Maritime traffic networks are a useful technique for compressing and abstracting multivessel trajectory data, but are also essential for the analysis of behavior of individual vessels or vessel groups and the detection of abnormal cases (Holst andEkman 2003, Holst et al 2012), such as search and rescue maneuvers (Varlamis et al 2018, Chatzikokolakis et al 2018, or other potentially illegal activities. The proposed work directly compares to works that extract maritime traffic models from multi-vessel AIS data (Andrienko and Andrienko 2011, Arguedas et al 2017, Coscia et al 2018. Arguedas et al (2017) propose a two-layer network (an external layer that contains the traffic network abstraction, and an internal layer that provides information about each vessel individually).…”

Section: Related Workmentioning

confidence: 99%

“…Arguedas et al (2017) propose a two-layer network (an external layer that contains the traffic network abstraction, and an internal layer that provides information about each vessel individually). Coscia et al (2018) create a graph-based representation of the maritime traffic, by extracting waypoints (nodes of the graph), indicating areas where vessels change course and by creating sea lanes (edges of the graph) between waypoints in the form of spatial ellipses. Pallotta et al (2013) propose a methodology for anomaly detection based on a maritime traffic model.…”

Section: Related Workmentioning

confidence: 99%

See 2 more Smart Citations

A distributed framework for extracting maritime traffic patterns

Kontopoulos

Varlamis

Tserpes

2020

International Journal of Geographical Information Science

View full text Add to dashboard Cite

All the modern surveillance systems take advantage of the Automatic Identification System (AIS), a compulsory tracking system for many types of vessels. Ships that carry AIS transponders on board transmit their position and status in order to alert nearby vessels and ground stations, but this information can well be used to identify events of interest and support decision making. The detection of anomalies (i.e. unexpected sailing behavior) in vessels' trajectories is such an event, which is of utmost importance. Approaches for detecting such anomalies vary from extracting normality models to searching for individual cases, such as AIS switch-off or collision avoidance maneuvers. The current research work follows the former method; it employs sparse historic AIS data and polynomial interpolation in order to extract shipping lanes. It modifies the DB-Scan clustering algorithm in order to achieve more coherent trajectory clusters, which are then composed to create the shipping lanes. The proposed approach implements distributed processing on Apache Spark in order to improve processing speed and scalability and is evaluated using real-world AIS data collected from terrestrial AIS receivers. The evaluation shows that the biggest part (i.e. more than 90%) of any future vessel trajectory falls within the extracted shipping lanes.

show abstract

Section: Resultsmentioning

confidence: 99%

Section: Related Workmentioning

confidence: 99%

Section: Related Workmentioning

confidence: 99%

See 1 more Smart Citation

A distributed framework for extracting maritime traffic patterns

Kontopoulos

Varlamis

Tserpes

2020

International Journal of Geographical Information Science

View full text Add to dashboard Cite

show abstract

“…Many works the recent years try to build maritime traffic network representations from historical AIS data [5,6]. Arguedas et al [5] propose a two-layer network: (i) an external layer that uses way-points as nodes/vertices and routes as edges/lines and (ii) an internal layer that consists of nodes or breakpoints that represent vessels' changes in behavior and edges or tracklets that represent vessel trajectories.…”

Section: Related Workmentioning

confidence: 99%

Uncovering Hidden Concepts from AIS Data: A Network Abstraction of Maritime Traffic for Anomaly Detection

Kontopoulos

Varlamis

Tserpes

2020

Lecture Notes in Computer Science

View full text Add to dashboard Cite

The compulsory use of Automatic Identification System (AIS) for many vessel types, which has been enforced by naval regulations, has opened new opportunities for maritime surveillance. AIS transponders are rich sources of information that everyone can collect using an RF receiver and provide real-time information about vessels' position. Properly taking advantage of AIS data, can uncover potential illegal behavior, offer real-time alerts and notify the authorities for any kind of anomalous vessel behavior. In this article, we extend an existing network abstraction of maritime traffic, that is based on nodes (called way-points) that correspond to naval areas of long stays or major turns for vessels (e.g. ports, capes, offshore platforms etc.) and edges (called traversals) that correspond to the routes followed by vessels between two consecutive way-points. The current work, focuses on the connections of this network abstraction and enriches them with semantic information about the different ways that vessels employ when traversing an edge. For achieving this, it proposes an alternative of the popular density based clustering algorithm DB-Scan, which modifies the proximity parameter (i.e. epsilon) of the algorithm. The proposed alternative employs in tandem the difference in (i) speed, (ii) course and (iii) position for defining the distance between two consecutive vessel positions (two consecutive AIS signals received from the same vessel). The results show that this combination performs significantly better than using only the spatial distance and, more importantly, results in clusters that have very interesting properties. The enriched network model can be processed and further examined with data mining techniques, even in an unsupervised manner, in order to identify anomalies in vessels' trajectories. Experimental results on a real dataset show the network's potential for detecting trajectory outliers and uncovering deviations on a vessel's route.

show abstract

“…More sophisticated methods suppose that vessel trajectories follow a distribution and learn it from historical data [5], [6]. Currently, state-of-the-art methods for trajectory reconstruction [7], [8], [9] use the following typical three-step approach: i) the first step involves a clustering method, e.g. TRACLUS [10] or TREAD [11] to cluster historical motion data into route patterns, ii) the second one assigns the vessel to be processed to one of these clusters iii) the third one interpolates or predicts the vessel trajectory based on the route pattern of the assigned cluster.…”

Section: Related Workmentioning

confidence: 99%

A Multi-Task Deep Learning Architecture for Maritime Surveillance Using AIS Data Streams

Nguyen

Vadaine

Hajduch

et al. 2018

2018 IEEE 5th International Conference on Data Science and Advanced Analytics (DSAA)

View full text Add to dashboard Cite

In a world of global trading, maritime safety, security and efficiency are crucial issues. We propose a multi-task deep learning framework for vessel monitoring using Automatic Identification System (AIS) data streams. We combine recurrent neural networks with latent variable modeling and an embedding of AIS messages to a new representation space to jointly address key issues to be dealt with when considering AIS data streams: massive amount of streaming data, noisy data and irregular timesampling. We demonstrate the relevance of the proposed deep learning framework on real AIS datasets for a three-task setting, namely trajectory reconstruction, anomaly detection and vessel type identification.

show abstract

Multiple Ornstein–Uhlenbeck Processes for Maritime Traffic Graph Representation

Cited by 62 publications

References 18 publications

A distributed framework for extracting maritime traffic patterns

A distributed framework for extracting maritime traffic patterns

Uncovering Hidden Concepts from AIS Data: A Network Abstraction of Maritime Traffic for Anomaly Detection

A Multi-Task Deep Learning Architecture for Maritime Surveillance Using AIS Data Streams

Contact Info

Product

Resources

About