Space-Based Global Maritime Surveillance. Part II: Artificial Intelligence and Data Fusion Techniques
Open accessThe DOI matches its Crossref record; 65 of the 71 references it cites were found in a registry or an index; 16 of the 55 sources its claims cite were checked for support.
Abstract
Maritime surveillance (MS) is of paramount importance for search and rescue operations, fishery monitoring, pollution control, law enforcement, migration monitoring, and national security policies. Since ground-based radars and automatic identification system (AIS) do not always provide a comprehensive and seamless coverage of the entire maritime domain, the use of space-based sensors is crucial to complement them. We reviewed space-based technologies for MS in the first part of this work, titled Space-based Global Maritime Surveillance. Part I: Satellite Technologies. However, MS systems combining multiple terrestrial and space-based sensors with additional information sources require dedicated artificial intelligence and data fusion techniques for processing raw satellite images and fusing heterogeneous information. The second part of our work focuses on some recent promising artificial intelligence and data fusion techniques for MS using space-based sensors.
Overview
CoChat Summary
TL;DR
This review surveys artificial intelligence, deep learning, and Bayesian/data-fusion techniques for space-based maritime surveillance, covering satellite image segmentation/classification, knowledge discovery from Sat-AIS, anomaly detection, and sum-product-algorithm-based multitarget tracking. It illustrates these methods with use cases on VHR/HR/Sentinel-2/COSMO-SkyMed imagery and AIS/SAR fusion, including an MTG extraction and a Malta-area SAR/AIS tracking scenario.
Key findings
- DNN use cases include Mask R-CNN trained on about 200,000 VHR images (≈1.5 m resolution, 768×768 pixels) from GeoEye, SPOT, Pleiades, and Black Sky for ship detection/feature extraction; U-Net on a synthetic HR dataset obtained by degrading 10,000 VHR SPOT images; ResNet34 for stern/bow classification on Sentinel-2 clipped ships; and an AIS-based vessel classifier with 2 hidden layers (200 and 100 neurons) and 14 output categories.
- Historical Sat-AIS knowledge discovery via maritime traffic graphs (MTGs) is demonstrated with 19,009 AIS tracks, 162,662 detected waypoints, 2,286 waypoint clusters, and a pruned graph of 99 nodes and 401 edges; the method was validated on real AIS datasets from the Iberian Coast and English Channel and in four MARISA operational trials.
- Maritime anomaly detection is formulated using the Ornstein-Uhlenbeck mean-reverting model with generalized likelihood ratio testing and joint Bayesian detection/tracking via hybrid Bernoulli filtering; it can detect deviations from standard routes and “dark” deviations during surveillance gaps, and was applied to a real-world rendezvous detection scenario.
- SPA-based multisensor MTT is presented as flexible and scalable for heterogeneous fusion, Sat-AIS integration, and class-aided tracking; a use case off Malta from 8 September 2018 at 16:30 to 12 September 2018 at 16:47 fused COSMO-SkyMed SAR measurements with AIS messages (AIS interval at least 5 hours) and identified a single target, the vehicles carrier GAIA LEADER, with real and estimated trajectories.
Methods
- Design: review/overview paper, companion to Part I, synthesizing AI/DNN and Bayesian/data-fusion techniques for maritime surveillance using space-based sensors.
- Data/sample: satellite images (VHR ≈1.5 m, HR ≈10 m, Sentinel-2, COSMO-SkyMed SAR), historical Sat-AIS data, AIS messages, and SAR measurements; use cases include training datasets of about 200,000 VHR images and 10,000 degraded VHR images.
- Main analysis: descriptive review and illustrative use cases rather than a controlled comparative experiment; quantitative results are limited mainly to dataset/graph counts and the Malta SAR/AIS tracking example.
Limitations
- The paper does not state an explicit limitations section. It notes Sat-AIS data challenges: erroneous, incomplete, intermittent, or counterfeit information; irregular sampling; coverage gaps from minutes to many hours; and intentional AIS disablement.
- It states limitations of existing MTT algorithms: fusion of heterogeneous information sources is not straightforward, they do not adapt to time-varying model parameters, and their complexity usually does not scale well with relevant system parameters such as the number of sensors.
- Evident from the text: as a review, the use cases are illustrative and not comprehensive; most DNN use cases report no accuracy or performance metrics, and the second HR segmentation use case relies on synthetic data because a real HR dataset was unavailable (the latter is stated in the paper).
Model-written summary. Check it against the paper before relying on it.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
9 Valid · 8 Not valid · 19 Not checked · 41 Own results and interpretations
The cited sources that could be checked couldn’t confirm or dispute this claim.
In the paper
“Complex DNNs with multiple hidden layers and neurons can be used for satellite image segmentation and target classification.”
What the sources say
- Could not verifycites [2] [2] I. Goodfellow, Y. Bengio, and A. Courville, 'Deep learning,' 2016, book in preparation for MIT Press. [Online]. Available: http://www.deeplearningbook.org - sourceExperimental — accuracy not yet measured
Our copy of Vivone et al. (2015) is entirely about knowledge-based multi-target ship tracking with HF surface-wave radar using a VS-IMM estimator combined with JPDA and UKF (Abstract p. 2; Index Terms p. 3; Introduction pp. 4-5; Conclusions p. 30). No passage mentions deep neural networks, hidden layers/neurons, satellite image segmentation, or learned target classification; searches on 'deep neural network hidden layers satellite image segmentation', 'neural network machine learning classifier' and 'image segmentation deep learning hidden layers neurons' returned only tracking/clutter passages (e.g. E. Adaptive Filter Module Selection, p. 13). The only satellite reference is a background remark that SAR satellites cannot give continuous coverage (Introduction, p. 3), which neither supports nor contradicts the DNN claim, so the claim cannot be checked against this source. (LLM, conf=0.75)
plainstates a cause- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Could not verify
The cited sources that could be checked couldn’t confirm or dispute this claim.
In the paper
“Space-based sensors for Earth observation installed on satellites allow collecting images of very large areas in remote regions of the globe with relatively short latency, and hence are strongly relevant to MS.”
What the sources say
- Could not verifycites [1] [1] G. Soldi, D. Gaglione, N. Forti, A. D. Simone, F. C. Daffin` a, G. Bottini, D. Quattrociocchi, L. M. Millefiori, P. Braca, S. Carniel, P. Willett, A. Iodice, D. Riccio, and A. Farina, 'Space-based global maritime surveillance. Part I: Satellite technologies,' IEEE Aerosp. Electron. Syst. Mag. (Submitted) , 2020. - sourceExperimental — accuracy not yet measured
The introduction (p. 1) discusses radar, sonar, and cameras as sensors for multitarget tracking, but does not address satellite Earth-observation imagery. Searches for satellite imagery, remote-area coverage, and latency found no passage supporting the attributed claim. (LLM, conf=0.76)
plainstates a cause- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Could not verify
The cited sources that could be checked couldn’t confirm or dispute this claim.
In the paper
“Besides the most common terrestrial sensors for MS, e.g., the automatic identification system (AIS), X-band radars, and over-the-horizon (OTH) radars, space-based sensor technologies enable persistent monitoring of the maritime domain and ship traffic on a global scale even in remote areas of the Earth.”
What the sources say
- Could not verifycites [1] [1] G. Soldi, D. Gaglione, N. Forti, A. D. Simone, F. C. Daffin` a, G. Bottini, D. Quattrociocchi, L. M. Millefiori, P. Braca, S. Carniel, P. Willett, A. Iodice, D. Riccio, and A. Farina, 'Space-based global maritime surveillance. Part I: Satellite technologies,' IEEE Aerosp. Electron. Syst. Mag. (Submitted) , 2020. - sourceExperimental — accuracy not yet measured
The introduction (p. 1) discusses multitarget tracking for surveillance using sensors such as radar, sonar, and cameras; the conclusion (p. 12) concerns the tracking algorithm. Searches for space-based sensors, maritime ship traffic, and global monitoring found no passage addressing the claimed persistent coverage of remote areas. (LLM, conf=0.76)
plainstates a cause- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Could not verify
The sources with a verdict support this claim; the rest weren’t checked yet, have no full text we could read, or their verdicts aren’t rated yet.
In the paper
“An emerging approach to MTT and information fusion - one with flexibility, low complexity and useful scalability - is based on a factor graph and the sum-product algorithm (SPA) [38], [39], [49].”
What the sources say
- Supportedcites [38] [38] F. Meyer, P. Braca, P. Willett, and F. Hlawatsch, 'A scalable algorithm for tracking an unknown number of targets using multiple sensors,' IEEE Trans. Signal Process. , vol. 65, no. 13, pp. 3478-3493, Jul. 2017. - source
The Proposed Method and Other Message Passing Methods (p. 2) describes multisensor multitarget tracking based on a factor graph and belief propagation using the sum-product algorithm. The same section reports low complexity, favorable scaling, and suitability for general nonlinear, non-Gaussian models, supporting the cited description of flexibility. (LLM, conf=0.94)
- Not checked — no full text of the sourcecites [39] [39] F. Meyer, T. Kropfreiter, J. L. Williams, R. A. Lau, F. Hlawatsch, P. Braca, and M. Z. Win, 'Message passing algorithms for scalable multitarget tracking,' Proc. IEEE , vol. 106, no. 2, pp. 221-259, Feb. 2018. - source
- Could not verifycites [49] [49] D. Gaglione, G. Soldi, F. Meyer, F. Hlawatsch, P. Braca, A. Farina, and M. Z. Win, 'Bayesian information fusion and multitarget tracking for maritime situational awareness,' IET Radar Sonar Navi. (in press) , 2020. - sourceExperimental — accuracy not yet measured
Section 8.1 describes vessel tracking by associating detections near predicted positions, but does not discuss a factor graph or the sum-product algorithm. Searches for those terms and for the claimed flexibility, low complexity, and scalability found no passage addressing this approach. (LLM, conf=0.76)
plainstates a cause- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Supported
The cited sources that could be checked couldn’t confirm or dispute this claim.
In the paper
“First, the fusion of heterogeneous information sources is not straightforward.”
What the sources say
- Could not verifycites [47] [47] A. Ponsford and J. Wang, 'A review of high frequency surface wave radar for detection and tracking of ships,' Turk. J. Elec. Eng. & Comp. Sci. , vol. 18, pp. 409-428, May 2010. - sourceExperimental — accuracy not yet measured
Section 8.1 describes associating radar detections into vessel tracks, while Section 5.3 discusses a trade-off in Doppler processing. Neither addresses the difficulty of fusing heterogeneous information sources in MTT algorithms. Searches for “heterogeneous,” “fusion,” and multi-target tracking across different sources found no passage making that point. (LLM, conf=0.76)
- Not checked — no full text of the sourcecites [48] [48] S. Maresca, P. Braca, J. Horstmann, and R. Grasso, 'Maritime surveillance using multiple high-frequency surface-wave radars,' IEEE Trans. Geosci. Remote Sens. , vol. 52, no. 8, pp. 5056-5071, Aug. 2014. - source
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Could not verify
34 more claims · 7 with their cited sources checked
Create a free account to see all 77 claims and to verify the 2 references still open.
- Probably not a substantial claim. 27
- It may state two separate claims as one. 16
- Borderline: one of the tests for a substantial claim is unclear. 1
Citations
Cites
Deep Residual Learning for Image Recognition
He, K., Zhang, X., Ren, S., & Sun, J. (2016) · 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2016.90 · Cited by 235,474 · cited 2×
“The type of architecture employed in this case is the U-Net, and The third use case is related to the classification between ship stern and bow by means of a ResNet34 [9] architecture.”
BackgroundU-Net: Convolutional Networks for Biomedical Image Segmentation
Ronneberger, O., Fischer, P., & Brox, T. (2015) · Lecture notes in computer science · doi:10.1007/978-3-319-24574-4_28 · Cited by 99,404 · cited 3×
“The type of architecture employed in this case is the U-Net, and The third use case is related to the classification between ship stern and bow by means of a ResNet34 [9] architecture.”
MethodHe, K., Gkioxari, G., Dollar, P., & Girshick, R. (2017) · 2017 IEEE International Conference on Computer Vision (ICCV) · doi:10.1109/iccv.2017.322 · Cited by 29,137
(1996) · Cited by 20,179
(2016) · Cited by 11,679
Estimation with Applications to Tracking and Navigation: Theory, Algorithms and Software
Bar‐Shalom, Y., Li, X. ‐., & Kirubarajan, T. (2002) · doi:10.1002/0471221279 · Cited by 6,782
Deep Residual Learning for Image Recognition
He, K., Zhang, X., Ren, S., & Sun, J. (2015) · arXiv · doi:10.48550/arxiv.1512.03385 · Cited by 4,196
He, K., Gkioxari, G., Dollar, P., & Girshick, R. (2020) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/tpami.2018.2844175 · Cited by 3,457
An algorithm for tracking multiple targets
Reid, D. (1979) · IEEE Transactions on Automatic Control · doi:10.1109/tac.1979.1102177 · Cited by 3,059
Statistical Multisource-Multitarget Information Fusion
Mahler, R. (2007) · Cited by 2,456
Multitarget bayes filtering via first-order multitarget moments
Mahler, R. P. S. (2003) · IEEE Transactions on Aerospace and Electronic Systems · doi:10.1109/taes.2003.1261119 · Cited by 2,235
PHD filters of higher order in target number
Mahler, R. (2007) · IEEE Transactions on Aerospace and Electronic Systems · doi:10.1109/taes.2007.4407475 · Cited by 1,020
The cardinality balanced multi-target multi-Bernoulli filter and its implementations
Vo, B.-T., Vo, B.-N., & Cantoni, A. (2009) · IEEE Transactions on Signal Processing · doi:10.1109/tsp.2008.2007924 · Cited by 851 · cited 1×
“Existing MTT algorithms can be broadly classified as “vectortype” algorithms, such as the joint probabilistic data association (JPDA) filter [36] and the multiple hypothesis tracker (MHT) [40], [41], and “set-type” algorithms, such as the (cardinalized) probability hypothesis density filter [42]–[45] and multi-Bernoul…”
MethodAnalytic Implementations of the Cardinalized Probability Hypothesis Density Filter
Vo, B.-T., Vo, B.-N., & Cantoni, A. (2007) · IEEE Transactions on Signal Processing · doi:10.1109/tsp.2007.894241 · Cited by 824 · cited 1×
“Existing MTT algorithms can be broadly classified as “vector-type” algorithms, such as the joint probabilistic data association (JPDA) filter [41] and the multiple hypothesis tracker (MHT) [42], [43], and “set-type” algorithms, such as the (cardinalized) probability hypothesis density filter [44]–[47], and multi-Berno…”
MethodKendall, A., Badrinarayanan, V., & Cipolla, R. (2017) · Procedings of the British Machine Vision Conference 2017 · doi:10.5244/c.31.57 · Cited by 784 · cited 1×
“Nowadays, the most used CNN architectures are the SegNet [3], the U-Net [4]–[6], and the Mask R-CNN [7].”
MethodPallotta, G., Vespe, M., & Bryan, K. (2013) · Entropy · doi:10.3390/e15062218 · Cited by 695 · cited 1×
“Finally, vector-based methods [9] use a chain representation of maritime traffic, considering sea routes as a set of links connecting waypoints.”
MethodMessage Passing Algorithms for Scalable Multitarget Tracking
Meyer, F., Kropfreiter, T., Williams, J. L., Lau, R., Hlawatsch, F., Braca, P., & Win, M. Z. (2018) · Proceedings of the IEEE · doi:10.1109/jproc.2018.2789427 · Cited by 355
Self-citationKendall, A., Badrinarayanan, V., & Cipolla, R. (2015) · arXiv · doi:10.48550/arxiv.1511.02680 · Cited by 292 · cited 1×
“Nowadays, the most used CNN architectures are the SegNet [3], the U-Net [4]–[6], and the Mask R-CNN [7].”
MethodA Scalable Algorithm for Tracking an Unknown Number of Targets Using Multiple Sensors
Meyer, F., Braca, P., Willett, P., & Hlawatsch, F. (2017) · IEEE Transactions on Signal Processing · doi:10.1109/tsp.2017.2688966 · Cited by 230 · cited 15×
“Bayesian inference on the DM indices n,k and the target states x n,k can be still be performed by running the SPA algorithm on a suitable devised factor graph.”
MethodInfluentialSelf-citationXiao, Z., Fu, X., Zhang, L., & Goh, R. S. M. (2020) · IEEE Transactions on Intelligent Transportation Systems · doi:10.1109/tits.2019.2908191 · Cited by 205 · cited 1×
“Knowledge discovery of maritime traffic patterns from AIS data can be broadly classified [18], [19] into statistics-, grid-, and vector-based methods.”
BackgroundMaritime Surveillance Using Multiple High-Frequency Surface-Wave Radars
Maresca, S., Braca, P., Horstmann, J., & Grasso, R. (2014) · IEEE Transactions on Geoscience and Remote Sensing · doi:10.1109/tgrs.2013.2286741 · Cited by 144 · cited 1×
“Algorithms of both types have been developed and evaluated [49], [50], and several limitations have been noted.”
BackgroundSelf-citationA Multi-Task Deep Learning Architecture for Maritime Surveillance Using AIS Data Streams
Nguyen, D., Vadaine, R., Hajduch, G., Garello, R., & Fablet, R. (2018) · 2018 IEEE 5th International Conference on Data Science and Advanced Analytics (DSAA) · doi:10.1109/dsaa.2018.00044 · Cited by 143
Maritime anomaly detection and threat assessment
Lane, R. O., Nevell, D. A., Hayward, S. D., & Beaney, T. W. (2010) · 2010 13th International Conference on Information Fusion · doi:10.1109/icif.2010.5711998 · Cited by 129 · cited 3×
“For decades, anomaly detection strategies have been explored and applied in maritime traffic monitoring [8], [14], [15], [29]–[33] in order to detect unexpected ship stops or unexpected changes in course, i.”
BackgroundContext-Enhanced Information Fusion: Boosting Real-World Performance with Domain Knowledge
(2016) · Advances in Computer Vision and Pattern Recognition · doi:10.1007/978-3-319-28971-7 · Cited by 124
Modeling vessel kinematics using a stochastic mean-reverting process for long-term prediction
Millefiori, L. M., Braca, P., Bryan, K., & Willett, P. (2016) · IEEE Transactions on Aerospace and Electronic Systems · doi:10.1109/taes.2016.150596 · Cited by 119 · cited 4×
“This method builds on recent advances in long-term vessel motion modeling [17], [25] whereby the dynamics of ships can be effectively described by a piecewise Ornstein–Uhlenbeck (OU) mean-reverting stochastic process.”
BackgroundMethodInfluentialSelf-citationPrediction oof Vessel Trajectories From AIS Data Via Sequence-To-Sequence Recurrent Neural Networks
Forti, N., Millefiori, L. M., Braca, P., & Willett, P. (2020) · IEEE International Conference on Acoustics Speech and Signal Processing · doi:10.1109/icassp40776.2020.9054421 · Cited by 119
Self-citationMaritime Traffic Probabilistic Forecasting Based on Vessels’ Waterway Patterns and Motion Behaviors
Xiao, Z., Ponnambalam, L., Fu, X., & Zhang, W. (2017) · IEEE Transactions on Intelligent Transportation Systems · doi:10.1109/tits.2017.2681810 · Cited by 110 · cited 1×
“Gridbased methods [6]–[8], [10] divide the area of interest into a spatial grid of indexed cells, each dealing with the static and dynamic properties of those vessels that passed through the specific element of the grid.”
MethodAssociative Learning of Vessel Motion Patterns for Maritime Situation Awareness
Bomberger, N., Rhodes, B., Seibert, M., & Waxman, A. (2006) · 2006 9th International Conference on Information Fusion · doi:10.1109/icif.2006.301661 · Cited by 103 · cited 2×
“Gridbased methods [6]–[8], [10] divide the area of interest into a spatial grid of indexed cells, each dealing with the static and dynamic properties of those vessels that passed through the specific element of the grid.”
BackgroundMethodAn algorithm for tracking multiple targets
Reid, D. B. (1978) · IEEE Conference on Decision and Control (CDC) · doi:10.1109/cdc.1978.268125 · Cited by 101
Space-Based Global Maritime Surveillance. Part I: Satellite Technologies
Soldi, G., Gaglione, D., Forti, N., Simone, A. D., Daffinà, F. C., Bottini, G., Quattrociocchi, D., Millefiori, L. M., Braca, P., Carniel, S., et al. (2021) · IEEE Aerospace and Electronic Systems Magazine · doi:10.1109/maes.2021.3070862 · Cited by 98 · cited 2×
“Space-based remote sensing technologies, reviewed in the first part of this work [1], include satellite-based AIS (Sat-AIS), synthetic aperture radar (SAR), multispectral (MSP) and hyperspectral (HSP) optical sensors, and global navigation satellite systems reflectometry (GNSS-R).”
BackgroundSelf-citationUnsupervised learning of maritime traffic patterns for anomaly detection
Vespe, M., Visentini, I., Bryan, K., & Braca, P. (2012) · 9th IET Data Fusion & Target Tracking Conference (DF&TT 2012): Algorithms & Applications · doi:10.1049/cp.2012.0414 · Cited by 97 · cited 2×
“Most research on maritime anomaly detection [8], [29]–[31] relies on two steps: i) knowledge discovery of maritime traffic patterns from historical data, and ii) anomaly detection via unsupervised learning.”
BackgroundSelf-citationMaritime Anomaly Detection using Gaussian Process Active Learning
Kowalska, K., & Peel, L. (2012) · Research Publications (Maastricht University) · Cited by 93
Rhodes, B. J., Bomberger, N. A., & Zandipour, M. (2007) · 2007 10th International Conference on Information Fusion · doi:10.1109/icif.2007.4408127 · Cited by 92
Self-Tuning Algorithms for Multisensor-Multitarget Tracking Using Belief Propagation
Soldi, G., Meyer, F., Braca, P., & Hlawatsch, F. (2019) · IEEE Transactions on Signal Processing · doi:10.1109/tsp.2019.2916764 · Cited by 89 · cited 1×
“A DM index ‘n;k is introduced for each target k to select the most appropriate DM at time n [53], [54].”
BackgroundSelf-citationKnowledge-Based Multitarget Ship Tracking for HF Surface Wave Radar Systems
Vivone, G., Braca, P., & Horstmann, J. (2015) · IEEE Transactions on Geoscience and Remote Sensing · doi:10.1109/tgrs.2014.2388355 · Cited by 88 · cited 1×
“MTT methods have to cope with various challenges, for example, the heterogeneity of the different information sources [38]–[40], and the measurement-origin uncertainty (MOU), i.”
Self-citationValsamis, A., Tserpes, K., Zissis, D., Anagnostopoulos, D., & Varvarigou, T. (2017) · Journal of Systems and Software · doi:10.1016/j.jss.2016.06.016 · Cited by 82
Multiple Ornstein–Uhlenbeck Processes for Maritime Traffic Graph Representation
Coscia, P., Braca, P., Millefiori, L. M., Palmieri, F. A. N., & Willett, P. (2018) · IEEE Transactions on Aerospace and Electronic Systems · doi:10.1109/taes.2018.2808098 · Cited by 77 · cited 3×
“Further advancements on vector-based knowledge discovery [11], [12] have recently led to the development of an unsupervised graph-based methodology to identify the spatiotemporal dynamics of ship routes, and efficiently extract a compact representation of global maritime patterns from large volumes of historical AIS d…”
BackgroundMethodSelf-citationDistributed Joint Attack Detection and Secure State Estimation
Forti, N., Battistelli, G., Chisci, L., Li, S., Wang, B., & Sinopoli, B. (2018) · IEEE Transactions on Signal and Information Processing over Networks · doi:10.1109/tsipn.2017.2760804 · Cited by 74
Self-citationDetecting anomalous deviations from standard maritime routes using the Ornstein-Uhlenbeck process
d'Afflisio, E., Braca, P., Millefiori, L. M., & Willett, P. (2018) · IEEE Transactions on Signal Processing · doi:10.1109/tsp.2018.2875887 · Cited by 59 · cited 2×
“More recently, an innovative statistical framework[32], [33] that combines the available context data with a parametric model of the vessel’s kinematic behavior has received special attention.”
BackgroundMethodSelf-citationA review of high frequency surface wave radar for detection and tracking of ships
PONSFORD, T., & WANG, J. (2010) · Turkish Journal of Electrical Engineering and Computer Sciences · doi:10.3906/elk-0912-331 · Cited by 58 · cited 1×
“Algorithms of both types have been developed and evaluated [49], [50], and several limitations have been noted.”
BackgroundRadar/AIS data fusion and SAR tasking for Maritime Surveillance
Guerriero, M., Willett, P., Coraluppi, S. P., & Carthel, C. A. (2008) · Cited by 56
Self-citationA distributed spatial method for modeling maritime routes
Zissis, D., Chatzikokolakis, K., Spiliopoulos, G., & Vodas, M. (2020) · IEEE Access · doi:10.1109/access.2020.2979612 · Cited by 53 · cited 1×
“Knowledge discovery of maritime traffic patterns from AIS data can be broadly classified [18], [19] into statistics-, grid-, and vector-based methods.”
BackgroundForty Years of Multiple Hypothesis Tracking - A Review of Key Developments
Chong, C.-Y., Mori, S., & Reid, D. B. (2018) · International Conference on Information Fusion (FUSION) · doi:10.23919/icif.2018.8455386 · Cited by 51 · cited 1×
“Existing MTT algorithms can be broadly classified as “vector-type” algorithms, such as the joint probabilistic data association (JPDA) filter [41] and the multiple hypothesis tracker (MHT) [42], [43], and “set-type” algorithms, such as the (cardinalized) probability hypothesis density filter [44]–[47], and multi-Berno…”
MethodBelief propagation based AIS/radar data fusion for multi-target tracking
Gaglione, D., Braca, P., & Soldi, G. (2018) · 2018 21st International Conference on Information Fusion (FUSION) · doi:10.23919/icif.2018.8455217 · Cited by 37 · cited 2×
“, ship’s position and the MMSI identifier [55], [56].”
BackgroundMethodSelf-citationBayesian information fusion and multitarget tracking for maritime situational awareness
Gaglione, D., Soldi, G., Meyer, F., Hlawatsch, F., Braca, P., Farina, A., & Win, M. Z. (2020) · IET Radar, Sonar & Navigation · doi:10.1049/iet-rsn.2019.0508 · Cited by 36 · cited 2×
“An emerging approach to MTT and information fusion— one with flexibility, low complexity, and useful scalability—is based on a factor graph and the SPA [39], [40], [51].”
MethodSelf-citationTrajectory compression-guided visualization of spatio-temporal AIS vessel density
Li, Y., Liu, R. W., Liu, J., Huang, Y., Hu, B., & Wang, K. (2016) · 2016 8th International Conference on Wireless Communications & Signal Processing (WCSP) · doi:10.1109/wcsp.2016.7752733 · Cited by 35 · cited 1×
“Common statistics-based techniques [15], [16] focus on analysing the available data in order to provide a visual representation and quantitative modelling of the fundamental statistical analytics.”
d'Afflisio, E., Braca, P., Millefiori, L. M., & Willett, P. (2018) · 2018 21st International Conference on Information Fusion (FUSION) · doi:10.23919/icif.2018.8455854 · Cited by 33 · cited 4×
“For decades, anomaly detection strategies have been explored and applied in maritime traffic monitoring [8], [14], [15], [29]–[33] in order to detect unexpected ship stops or unexpected changes in course, i.”
BackgroundMethodInfluentialSelf-citationUnsupervised extraction of maritime patterns of life from Automatic Identification System data
Forti, N., Millefiori, L. M., & Braca, P. (2019) · OCEANS 2019 - Marseille · doi:10.1109/oceanse.2019.8867429 · Cited by 33 · cited 4×
“Further advancements on vector-based knowledge discovery [11], [12] have recently led to the development of an unsupervised graph-based methodology to identify the spatiotemporal dynamics of ship routes, and efficiently extract a compact representation of global maritime patterns from large volumes of historical AIS d…”
BackgroundMethodInfluentialSelf-citationMillefiori, L. M., Braca, P., & Willett, P. (2016) · IEEE Signal Processing Letters · doi:10.1109/lsp.2016.2605705 · Cited by 30 · cited 1×
“This method builds on recent advances in long-term vessel motion modeling [17], [25] whereby the dynamics of ships can be effectively described by a piecewise Ornstein–Uhlenbeck (OU) mean-reverting stochastic process.”
MethodSelf-citationAnomaly Detection and Tracking Based on Mean–Reverting Processes with Unknown Parameters
Forti, N., Millefiori, L. M., Braca, P., & Willett, P. (2019) · ICASSP 2019 - 2019 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) · doi:10.1109/icassp.2019.8683428 · Cited by 29 · cited 1×
“An adaptive version [35] of this Bayesian framework is capable of also handling unknown parameters in the underlying OU vessel dynamics.”
BackgroundSelf-citationJoint attack detection and secure state estimation of cyber‐physical systems
Forti, N., Battistelli, G., Chisci, L., & Sinopoli, B. (2020) · International Journal of Robust and Nonlinear Control · doi:10.1002/rnc.4724 · Cited by 29 · cited 1×
“Based on recent advances on hybrid Bernoulli filtering [37], which combines standard Figure 6.”
MethodSelf-citationData-driven detection and context-based classification of maritime anomalies
Pallotta, G., & Jousselme, A.-L. (2015) · Cited by 29
Historical Handwritten Document Segmentation by Using a Weighted Loss
Capobianco, S., Scommegna, L., & Marinai, S. (2018) · Lecture notes in computer science · doi:10.1007/978-3-319-99978-4_31 · Cited by 19
Classification-Aided Multitarget Tracking Using the Sum-Product Algorithm
Gaglione, D., Soldi, G., Braca, P., Magistris, G. D., Meyer, F., & Hlawatsch, F. (2020) · IEEE Signal Processing Letters · doi:10.1109/lsp.2020.3024858 · Cited by 18 · cited 1×
“As a consequence, the inclusion of class information greatly improves the performance of the SPA-based MTT algorithm [57], [58].”
MethodSelf-citationSoldi, G., & Braca, P. (2018) · International Conference on Information Fusion (FUSION) · doi:10.23919/icif.2018.8455612 · Cited by 15 · cited 1×
“A DM index ‘n;k is introduced for each target k to select the most appropriate DM at time n [53], [54].”
BackgroundSelf-citationHybrid Bernoulli Filtering for Detection and Tracking of Anomalous Path Deviations
Forti, N., Millefiori, L. M., & Braca, P. (2018) · 2018 21st International Conference on Information Fusion (FUSION) · doi:10.23919/icif.2018.8455567 · Cited by 14 · cited 2×
“Further developments [34]–[36] focused on the problem of dynamic detection of motion anomalies and joint tracking for real-time MS.”
BackgroundMethodSelf-citationValidation of the Ornstein-Uhlenbeck route propagation model in the Mediterranean Sea
Millefiori, L. M., Pallotta, G., Braca, P., Horn, S., & Bryan, K. (2015) · OCEANS 2015 - Genova · doi:10.1109/oceans-genova.2015.7271565 · Cited by 13 · cited 1×
“This approach, extensively validated against real-world datasets [17], [26], relies on model parameters that change only at waypoints (places where ships regularly stop or change their velocity).”
BackgroundSelf-citationRandom finite set tracking for anomaly detection in the presence of clutter
Forti, N., Millefiori, L. M., Braca, P., & Willett, P. (2020) · 2020 IEEE Radar Conference (RadarConf20) · doi:10.1109/radarconf2043947.2020.9266705 · Cited by 11 · cited 2×
“Further developments [34]–[36] focused on the problem of dynamic detection of motion anomalies and joint tracking for real-time MS.”
BackgroundMethodSelf-citationHeterogeneous information fusion for multitarget tracking using the sum-product algorithm
Soldi, G., Gaglione, D., Meyer, F., Hlawatsch, F., Braca, P., Farina, A., & Z. Win, M. (2019) · ICASSP 2019 - 2019 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) · doi:10.1109/icassp.2019.8683891 · Cited by 10 · cited 2×
“, ship’s position and the MMSI identifier [55], [56].”
BackgroundMethodSelf-citationScalable distributed change detection and its application to maritime traffic
Millefiori, L. M., Braca, P., & Arcieri, G. (2017) · 2017 IEEE International Conference on Big Data (Big Data) · doi:10.1109/bigdata.2017.8258101 · Cited by 10 · cited 1×
“Based on the OU dynamic model (DM) for accurate longterm ship prediction, change detection algorithms can be applied to identify specific geospatial waypoints where the mean long-term velocity parameter of the underlying OU process tends to change [11], [27].”
MethodSelf-citationBehavioral learning of vessel types with fuzzy-rough decision trees
Falcón, R., Abielmona, R., & Blasch, E. (2014) · Cited by 10
Soldi, G., Gaglione, D., De Magistris, G., Braca, P., Stinco, P., Ferri, G., Tesei, A., & Le Page, K. (2020) · ICASSP 2020 - 2020 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) · doi:10.1109/icassp40776.2020.9054038 · Cited by 8 · cited 1×
“As a consequence, the inclusion of class information greatly improves the performance of the SPA-based MTT algorithm [55], [56].”
MethodSelf-citationAutomated port traffic statistics: From raw data to visualisation
Cazzanti, L., Davoli, A., & Millefiori, L. M. (2016) · 2016 IEEE International Conference on Big Data (Big Data) · doi:10.1109/bigdata.2016.7840765 · Cited by 6
Self-citationSemantic Segmentation using Deep Learning: A case of study in Albufera Park, Valencia
Gargiulo, M., Dell’Aglio, D. A. G., Iodice, A., Riccio, D., & Ruello, G. (2019) · IEEE International Workshop on Metrology for Agriculture and Forestry (MetroAgriFor) · doi:10.1109/metroagrifor.2019.8909243 · Cited by 5 · cited 1×
“Nowadays, the most used CNN architectures are the SegNet [3], the U-Net [4]–[6], and the Mask R-CNN [7].”
MethodSelf-citation(2008)
Research works printed but not linked: 1
- [8] E. H. Adelson, C. H. Anderson, J. R. Bergen, P. J. Burt, and J. M. Ogden, 'Pyramid methods in image processing,' RCA Engineer , vol. 29, no. 6, pp. 33-41, 1984. -
Cited by
Artificial Intelligence for Trusted Autonomous Satellite Operations
Thangavel, K., Sabatini, R., Gardi, A., Ranasinghe, K. C., Hilton, S., Servidia, P. A., & Spiller, D. (2023) · Progress in Aerospace Sciences · doi:10.1016/j.paerosci.2023.100960 · Cited by 144
Nguyen, D., & Fablet, R. (2024) · IEEE Access · doi:10.1109/access.2024.3349957 · Cited by 215
“…time horizons, ranging from a few tens of minutes to tens of hours ahead, is at the core of diverse MS and MSA applications, including but not limited to search and rescue [1], [2], traffic control [3], path planning [4], [5], [6], port congestion avoidance [7], [8], [9], pollution monitoring [10].”
BackgroundSpace-Based Global Maritime Surveillance. Part I: Satellite Technologies
Soldi, G., Gaglione, D., Forti, N., Simone, A. D., Daffinà, F. C., Bottini, G., Quattrociocchi, D., Millefiori, L. M., Braca, P., Carniel, S., et al. (2021) · IEEE Aerospace and Electronic Systems Magazine · doi:10.1109/maes.2021.3070862 · Cited by 98
“The second part of this work [1] focuses on artificial intelligence and”
BackgroundInfluentialSelf-citationZhang, Z., Zhang, L., Wu, J., & Guo, W. (2024) · IEEE Geoscience and Remote Sensing Magazine · doi:10.1109/mgrs.2024.3404506 · Cited by 112
The usefulness of artificial intelligence for safety assessment of different transport modes
Tselentis, D. I., Papadimitriou, E., & van Gelder, P. H. A. J. M. (2023) · Accident Analysis & Prevention · doi:10.1016/j.aap.2023.107034 · Cited by 99
Metadata
- Authors
- Soldi, G., Gaglione, D., Forti, N., Millefiori, L. M., Braca, P., Carniel, S., Simone, A. D., Iodice, A., Riccio, D., Daffina, F. C., Quattrociocchi, D., Bottini, G., Willett, P., & Farina, A.
- Affiliations
- NATO Centre for Maritime Research and Experimentation (IT)
- University of Naples Federico II (IT)
- e GEOS (Italy) (IT)
- Food and Agriculture Organization of the United Nations (IT)
- University of Connecticut (US)
- ORCID
- Venue
- IEEE Aerospace and Electronic Systems Magazine
- Published
- 2021-09-01
- Volume
- 36
- Issue
- 9
- Pages
- 30-42
- ISSN
- 0885-8985 (ISSN-L)
- 1557-959X
- OpenAlex
- W3110443977
- Cited by
- 44
- References
- 71
- Citations by year
- 2026: 6 · 2025: 7 · 2024: 11 · 2023: 6 · 2022: 7 · 2021: 7
- FWCI
- 4.11
- Type
- journal article
- Language
- en
- Open access
- green
- arxiv.org
- Licence
- Locations
- IEEE Aerospace and Electronic Systems Magazine (publishedVersion)
- arXiv (Cornell University) (submittedVersion)
- CINECA IRIS Institutial research information system (University of Pisa) (submittedVersion)
- Florence Research (University of Florence) (submittedVersion)
- arXiv (Cornell University)
- Funders
- NATO Allied Command Transformation (SAC000A08)
- NIUVT
- Air Force Office of Scientific Research (FA9500-18-1-0463)
- Topics
- Physical Sciences › Engineering › Ocean Engineering › Maritime Navigation and Safety
- Physical Sciences › Engineering › Media Technology › Advanced Image Fusion Techniques
- Physical Sciences › Computer Science › Artificial Intelligence › Target Tracking and Data Fusion in Sensor Networks
- Keywords
- maritime surveillance, satellite sensors, satellite imagery, artificial intelligence, data fusion, heterogeneous data fusion, search and rescue, fisheries monitoring, pollution monitoring, migration monitoring, national security, Automatic Identification System
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
The claim verdicts this report counts are experimental: their accuracy is not yet measured.
A language model drew the claims from the paper on 2026-10-10, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).
A second model labelled each claim on 2026-10-10 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).
Cited sources are checked with a language model’s help (qwen3.8-omni-flash and gpt-6-sol).
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.