Space-Based Global Maritime Surveillance. Part II: Artificial Intelligence and Data Fusion Techniques

Soldi, Giovanni; Gaglione, Domenico; Forti, Nicola; Millefiori, Leonardo M.; Braca, Paolo; Carniel, Sandro; Simone, Alessio Di; Iodice, Antonio; Riccio, Daniele; Daffina, Filippo Cristian; Quattrociocchi, Dino; Bottini, Gianfausto; Willett, Peter; Farina, A.

doi:10.1109/maes.2021.3070884

Cited by 36 publications

(12 citation statements)

References 45 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…More recently, the increasing amount of free and open optical data, e.g., S-2, Landsat 8, along with the advances in hardware (e.g., GPU computing), has boosted the application of machine learning paradigms, e.g., deep learning, convolutional neural networks and support-vector machines, to both detection and classification purposes for MS (see further details in the companion paper [1]). To summarize, a high level of spatial detail along with the easier interpretability and the absence of speckle noise make optical satellite imagery a very attractive solution for maritime surveillance purposes.…”

Section: Msp Optical Sensorsmentioning

confidence: 99%

“…To summarize, 2020/11/24 DRAFT Note that, the availability of multiple space-based sensors, providing images of the current maritime situational picture at different spectral and spatial resolutions, requires also the development of advanced artificial intelligence and data fusion algorithms. These techniques have to cope with the processing and fusion of valuable information acquired from multiple sensors and will be covered in the second part of this work [1].…”

Section: Gnss-rmentioning

confidence: 99%

“…This paper is divided into two parts: the first part provides an overview of the main available space-based sensors technologies, i.e., Sat-AIS, SAR, MSP and HSP optical sensors, and GNSS-R and presents the advantages and limitations of each technology in the scope of MS. The second part of this work[1] focuses on artificial intelligence and data fusion techniques to extract information from raw satellite data and fuse this information to extract valuable and actionable knowledge for end-users and decision-makers.II. SATELLITE TECHNOLOGIES FOR MARITIME SURVEILLANCESpace-based sensors are usually installed on satellites which orbit around the Earth and they can serve different purposes.…”

mentioning

confidence: 99%

See 2 more Smart Citations

Space-Based Global Maritime Surveillance. Part I: Satellite Technologies

Soldi

Gaglione

Forti

et al. 2021

IEEE Aerosp. Electron. Syst. Mag.

Self Cite

View full text Add to dashboard Cite

Maritime surveillance (MS) is crucial for search and rescue operations, fishery monitoring, pollution control, law enforcement, migration monitoring, and national security policies. Since the early days of seafaring, MS has been a critical task for providing security in human coexistence. Several generations of sensors providing detailed maritime information have become available for large offshore areas in real time: maritime radar sensors in the 1950s and the automatic identification system (AIS) in the 1990s among them. However, ground-based maritime radars and AIS data do not always provide a comprehensive and seamless coverage of the entire maritime space. Therefore, the exploitation of spacebased sensor technologies installed on satellites orbiting around the Earth, such as satellite AIS data, synthetic aperture radar, optical sensors, and global navigation satellite systems reflectometry, becomes crucial for MS and to complement the existing terrestrial technologies. In the first part of this work, we provide an overview of the main available space-based sensors technologies and present the advantages and limitations of each technology in the scope of MS. The second part, related to artificial intelligence, signal processing and data fusion techniques, is provided in a companion paper, titled: "Space-based

show abstract

Section: Msp Optical Sensorsmentioning

confidence: 99%

Section: Gnss-rmentioning

confidence: 99%

mentioning

confidence: 99%

See 1 more Smart Citation

Space-Based Global Maritime Surveillance. Part I: Satellite Technologies

Soldi

Gaglione

Forti

et al. 2021

IEEE Aerosp. Electron. Syst. Mag.

Self Cite

View full text Add to dashboard Cite

show abstract

“…Although traditional satellite remote sensing already plays a role in monitoring illegal activity at sea (Kurekin et al., 2019; Oozeki et al., 2017), the detection of small wooden vessels often involved in IUU fishing (Collins et al., 2021) remains particularly difficult, especially considering the infrequent acquisitions of medium‐high resolution earth observation data over the oceans. A SmallSat constellation targeted on specific water masses (e.g., a marine protected area) with greater spatial resolution could better detect small vessels operating illegally (Kanjir et al., 2016; Lazreg et al., 2018), especially if combined with artificial intelligence (Soldi et al., 2020). These could be further coupled with animal‐borne devices detecting IUU (Weimerskirch et al., 2018, 2020), with SmallSats acting as a relay to transmit the data to managers in a timely manner.…”

Section: Opportunities Associated With Smallsatsmentioning

confidence: 99%

SmallSats: a new technological frontier in ecology and conservation?

Curnick

Davies²,

Duncan

et al. 2021

Remote Sens Ecol Conserv

View full text Add to dashboard Cite

We are in the midst of a revolution in satellite technology, with the rapid development and advancement of small satellites (or SmallSats, i.e., satellites <180 kg). Here, we review the opportunities and challenges that such technology might afford in the field of conservation and ecology. SmallSat constellations may yield higher resolutions than those that are currently available to scientists and practitioners, increasing opportunities to improve environmentalmonitoring and animal-tracking capabilities. They may cut access costs to end users, by reducing operational costs and bringing increased competition to the existing market. Their greater flexibility and affordability may moreover enable the development of bespoke constellations for specific conservation and ecological applications, and provide greater interoperability with ground-based sensors, such as tracking devices and camera traps. In addition, SmallSats may serve as cost-effective research and development platforms for new components and products. Combined, these benefits could significantly improve our ability to monitor threats to the environment as they unfold, while enhancing our understanding of animal ecology and ecosystem dynamics. However, significant hardware and software developments are required before such technology is able to produce, process and handle reliable and cost-effective data, and the initial research and development costs still represent a major challenge. Further, we argue that much remains to be done to ensure these new data products become accessible, equitable and sustainable.

show abstract

“…DL and deep RL are also key components of new-generation autonomous driving systems, see, e.g., [17], [18], and nowadays are also widely exploited in surveillance systems, such as Synthetic Aperture Radar (SAR) imaging, see, e.g., [19]- [22]. In space-based surveillance, DL offers the capability to accurately classify vessels from satellite sensor imaging (see, e.g., [23], [24]), and in the context of maritime situational awareness RNNs are able to accurately predict vessel positions several hours ahead [25]- [28]. In video analysis and image understanding, DL methods represent the state of the art for object detection [29] and multi-object tracking [30].…”

Section: Introductionmentioning

confidence: 99%

Statistical Hypothesis Testing Based on Machine Learning: Large Deviations Analysis

Braca¹,

Millefiori²,

Aubry³

et al. 2022

Preprint

Self Cite

View full text Add to dashboard Cite

We study the performance-and specifically the rate at which the error probability converges to zero-of Machine Learning (ML) classification techniques. Leveraging the theory of large deviations, we provide the mathematical conditions for a ML classifier to exhibit error probabilities that vanish exponentially, say ∼ exp (−n I + o(n)), where n is the number of informative observations available for testing (or another relevant parameter, such as the size of the target in an image) and I is the error rate. Such conditions depend on the Fenchel-Legendre transform of the cumulant-generating function of the Data-Driven Decision Function (D3F, i.e., what is thresholded before the final binary decision is made) learned in the training phase. As such, the D3F and, consequently, the related error rate I, depend on the given training set, which is assumed of finite size. Interestingly, these conditions can be verified and tested numerically exploiting the available dataset, or a synthetic dataset, generated according to the available information on the underlying statistical model. In other words, the classification error probability convergence to zero and its rate can be computed on a portion of the dataset available for training. Coherently with the large deviations theory, we can also establish the convergence, for n large enough, of the normalized D3F statistic to a Gaussian distribution. This property is exploited to set a desired asymptotic false alarm probability, which empirically turns out to be accurate even for quite realistic values of n. Furthermore, approximate error probability curves ∼ ζn exp (−n I) are provided, thanks to the refined asymptotic derivation (often referred to as exact asymptotics), where ζn represents the most representative sub-exponential terms of the error probabilities. Leveraging the refined asymptotic, we are able to compute an analytically accurate approximation of the classification performance for both the regimes of small and large values of n. Theoretical findings are corroborated by extensive numerical simulations. Real-world data, acquired by an X-band maritime radar system for surveillance, are exploited to assess the validity of the proposed theory.

show abstract

Space-Based Global Maritime Surveillance. Part II: Artificial Intelligence and Data Fusion Techniques

Cited by 36 publications

References 45 publications

Space-Based Global Maritime Surveillance. Part I: Satellite Technologies

Space-Based Global Maritime Surveillance. Part I: Satellite Technologies

SmallSats: a new technological frontier in ecology and conservation?

Statistical Hypothesis Testing Based on Machine Learning: Large Deviations Analysis

Contact Info

Product

Resources

About