Artificial Intelligence (AI) applications on Earth Observation (EO) satellite data, such as those for vessel detection, are gaining attention for their potential to meet strict bandwidth and latency requirements. While traditional on-ground computing pipelines often rely on heavy post-processing, implementing these techniques onboard satellites is challenging due to limited computing resources. To support the development of efficient onboard data processing strategies, this study compares the performance of object detection on raw data from Sentinel-2 and VENμS missions. The study demonstrates that the proposed two-stage approach with a focus on efficiency is capable of identifying vessels in raw data with minimal pre-processing. Specifically, our method achieved a remarkable Average Precision (AP) of 0.841 on the VENμS dataset.

Hyperspectral sensors are increasing in popularity for Earth Observation applications due to their ability to gather data over multiple spectral bands. However, the processing of such amount of information is difficult to handle for the current computing capabilities of small satellites. Several Band Selection methodologies have been developed in the last years; although, some of them demand very low computational resources, they use, at least, Level 1 data products. Therefore, the Level 0 data needs to be processed and the spectral bands coregistered. Artificial Intelligence has shown its potential to reduce the computational burden while achieving high accuracies in EO applications. In this study, a Neural Network-based methodology is proposed to select a spectral band set directly using non coregistered data captured by hyperspectral sensors.

This paper presents the findings of the COAST project, an innovative investigation into maritime surveillance using Synthetic Aperture Radar (SAR) technology, funded by the Italian Space Agency. The research specifically delves into the synergistic potential of a multi-frequency/multi-mission (MFMM) approach to detect and identify vessels, a facet relatively unexplored in previous studies. Four technical objectives drive the study, including the detection of both visible and non-visible non-cooperating ships and their classification using SAR tomography techniques, as well as AIS/SAR data matching with velocity estimation. This research promises to enhance maritime domain awareness and security by furthering our understanding of SAR-based vessel detection and classification methodologies.

Nowadays, the use of Artificial Intelligence on board Earth Observation satellites is under investigation for applications having strict bandwidth and latency requirements, such as vessel detection. However, many of the on-ground current computing pipelines rely on data post-processing techniques whose applications onboard satellites are tricky because of their limited computing power. To enable the analysis and the research of lightweight onboard data processing techniques, we provide VDS2Raw, the first Sentinel-2 Raw dataset for vessel detection applications. Finally, we also compared different object detection Deep Learning techniques in terms of detection performance and inference time to perform a feasibility analysis of performing onboard vessel detection on raw multi-spectral data.

The wakes generated by moving vessels represent relevant patterns in remotely sensed images. They are a marker of ship presence and can be processed to infer route, speed, size, and type of ships. Automatic wake detection can be exploited by law enforcement agencies and local authorities for ensuring a wide range of applications, including maritime traffic surveillance, border control, and protection of marine protected areas. The topic is thus attracting increasing interest from the remote sensing community. This paper contributes in this context presenting a novel approach based on the detection of the keypoints of wake components by Convolutional Neural Networks (CNNs) in electro-optical satellite imagery. The selected approach to deep learning relies on a transfer learning procedure fine-tuning the ImageNet weights. This is performed through an ad-hoc developed dataset realized from Sentinel-2 multi-spectral images and Automatic Identification System (AIS) data in northern Europe. The experimental results confirm the robustness of the proposed method, which is tested against different spectral bands from visible to near-infrared and also by a domain shifting on lower resolution Landsat-9 images. Fractional errors in the positioning of the wake vertex are lower than 10% and the achieved heading accuracy is below 10°. The proposed method is faster than the traditional approaches based on Radon Transform, and due to its lightweight nature, our model can be executed efficiently on edge-AI devices, enabling real-time processing onboard.

Maritime trade and trasport occupy a pivotal position in the current era of globalization. Thus, monitoring ships at sea represents the starting point of this paper in which a novel approach to detect ships by wake has been proposed, based on Instance Segmentation deep learning architecture Mask R-CNN. In order to train and test this network, 766 wake chips cropped from 50 multispectral images acquired from Sentinel-2 satellites were observed. In particular, B2 (blue), B3 (green), B4 (red) and B8 (Infrared) bands were considered since they are all characterized by same resolution. The results proved that Mask R-CNN is capable to detect the vast majority of ship wakes with high confidence percentage in different configurations, i.e. slanted wakes, multiple wake scenarios or wakes in dark areas not related to their features.

The recent launches of Earth Observation (EO) satellites have made numerous SAR images available to dynamically monitor the ocean with improved spatial resolution at shorter revisit time. Focusing on automatic target detection of ships with the specific aim of improving our Maritime Domain Awareness (MDA), this work assesses the capabilities of multi-frequency/multi-mission spaceborne Synthetic Aperture Radar (SAR) data. Specifically, Sentinel-1 (C-band), COSMO-SkyMed (X-band), and SAOCOM (L-band) missions have been considered in this analysis. The aim of the paper is to present an efficient approach for interpreting images acquired within small time gaps, ensuring fisheries and pollution control, anti-piracy actions, and surveillance over coastal/protected regions.

The paper shows the strong potentialities of multimission/multifrequency SAR data for improving the maritime situational awareness in coastal areas. Two main issues are analyzed: the detection of ships that are visible in SAR images and the identification of non-collaborative vessels, which are not visible in SAR images. In the first case, the multimission/multifrequency data guarantees: (a) smaller revisit time with respect to a single mission, enabling cross-check of the detection in several images and, thus, improving the detection rate, and (b) the availability of images covering large areas at low resolution as well as smaller swath observed with higher resolution. This is crucial in particular for the coastal areas where local phenomena can strongly affect the detection performance. In the second case, the multimission/multifrequency data enables innovative approaches exploiting the different appearance of ship and its wake at different frequencies.