Maritime situation awareness is supported by a combination of satellite, airborne, and terrestrial sensor systems. This paper presents several solutions to process that sensor data into information that supports operator decisions. Examples are vessel detection algorithms based on multispectral image techniques in combination with background subtraction, feature extraction techniques that estimate the vessel length to support vessel classification, and data fusion techniques to combine image based information, detections from coastal radar, and reports from cooperative systems such as (satellite) AIS. Other processing solutions include persistent tracking techniques that go beyond kinematic tracking, and include environmental information from navigation charts, and if available, ELINT reports. And finally rule-based and statistical solutions for the behavioural analysis of anomalous vessels. With that, trends and future work will be presented.

Because of global economic and socio-political changes, an increase of conflicts near the world's coastlines is anticipated. The littoral zone is characterized by intense regular vessel traffic. The conduct of Maritime Security Operations and Peace support Operations therefore means that navies have to control instead of dominate the sea, allowing regular vessel traffic in the area of operations, and act against irregular adversaries who nevertheless also can possess military armaments. Piracy, drug trafficking and other threatening events become obscured in the crowd of everyday fisheries, cargo traders, ferries and pleasure cruises, hindering the detection of anomalies and suspect behavior, and resulting in insufficient situation awareness. For controlling these situations information superiority and adequate situation awareness is a necessity. For the purpose to achieve information superiority a research program at TNO, in collaboration with the RNLN, has started aiming at improving maritime situation awareness. To improve situation awareness and threat detection capabilities in maritime scenarios the combination of sensor-based information with context information and intelligence from various sources is required. In the study the fusion and analysis for revealing anomalies and suspect from normal behavior are based on domain ontologies. A test bed allows the study of various exploitation and assessments techniques applied to these domain ontologies. Using an appropri- ate scenario we have simulated suspect and normal behaviour to test the applicability of these techniques.

About one year of coastal AIS (Automatic Identification System) data are analysed for vessel behaviour. The data are obtained from thousands of vessels along the Dutch North Sea coast including the entrance to the main port of Rotterdam. Vessel behaviour is described by a number of basic indicators that can be obtained directly from the AIS messages or from other sensors like radar. It is investigated here if the indicators may be used in e.g. classification algorithms to detect suspicious vessels for the purpose of maritime and border security. In addition, the AIS data are used for a preliminary assessment of the vessel detection possibilities of the recently launched TerraSAR-X satellite when operating in ScanSAR mode.

During the last years, the demand for vessel surveillance has increased both for fisheries control and for maritime security and safety. In order to overcome the limitations posed by conventional systems, surveillance with satellite SAR is being adopted more frequently because of its possibility to provide ship detection over wide swaths and under many conditions. Up to now, different satellite sensors have been available for vessel surveillance. Nevertheless, considering revisit and coverage requirements, the availability of additional satellite sensors is of interest. Thus, this study evaluates the quality of the new TerraSAR-X (ScanSAR mode) and gives a preliminary impression about its performance for vessel detection. For the study, two images were acquired over The Netherlands’ Western sea board in April 2008. Concurrent information on vessel positions was available from AIS and (limited) VMS. On both the acquisitions, weather and sea conditions were quite calm, resulting in very low background sea clutter enabling recognition of quite weak features. The two ScanSAR images have a resolution of 16 m, incidence angles around 39 degrees and HH polarisation. They were analysed with automatic ship detection software (“SUMO”) and also manually inspected. The images contain many ships and also platforms and windmills. Generally speaking, ship detection is limited by noise, clutter, in-homogeneities, features and artefacts in the image background, which hamper detection of the weakest targets or cause false alarms. Specifically these images show, concerning natural features on the sea surface: “cloud”-like wind inhomogeneities; sharp and more diffuse sea water fronts; shoals, banks and possibly some sea bottom topography; smallscale striations perpendicular to the coast; and waves breaking on the shore. Concerning image quality, a critical evaluation shows: striping parallel to the range direction with a period of about 2.5 km, and additionally as 1-pixel wide lines; azimuth ambiguities of strong targets at around 5 km offset; in one image, also range ambiguities at about 57 km offset; and km-long sidelobes in range and azimuth from a few very bright ships. The measured ENL is around 5-6 except in some bands parallel to the azimuth direction where it reaches 9-10; while this shows that optimum use has been made of the available raw data, such a variation of ENL within the image is undesirable for some automatic ship detection algorithms that work with a pre-defined ENL value. When comparing image geolocation, only using data in the main image tiff file, to GCPs in Google Earth, inaccuracies of up to 700 m are found. Although the effects mentioned above lead to false alarms (or, equivalently, necessitate a higher threshold setting leading to less detection sensitivity), the ships present in the images appear mostly with good contrast and, for the larger ones, welldefined outlines. In many cases wakes of the turbulent and Kelvin type are associated. In case of the AIS-carrying – larger – vessels, there is mostly no difficulty in detecting them, and the length and heading estimated from the SAR image compares mostly well with the AIS values. For the VMS-equipped fishing vessels, 4 out of 14 vessels with a length below 20 m could be seen. Based on this limited sample, as preliminary conclusion it can be said that the TerraSAR-X ScanSAR images are well suitable for vessel detection over wide areas in calm conditions.

PHARUS is a polarimetric phased array C-band Synthetic Aperture Radar (SAR), designed and built for airborne use. Advanced SAR (ASAR) data in image and alternating polarization mode have been simulated with PHARUS to demonstrate the use of Envisat for a number of typical SAR applications that are not possible with the ERS satellites. These applications concern landuse classification with several combinations of polarimetric channels and terrain height estimation from a same side stereo pair. In addition, SAR image geocoding and the determination of directional ocean wave spectra have been investigated as well. PHARUS data collected over the Black Forest in Germany, over a cultivated area in The Netherlands, and over the North Sea, were degenerated in the processing to ASAR resolution with two looks in alternating polarization mode, four looks in image mode, and one look in wave mode. Results show that ASAR dual-channel data from alternating polarization mode can better separate vegetation types than the single-channel data of the ERS satellites. Moreover, with ASAR resolution, terrain relief can be estimated with an accuracy of about 50 m, whereas SAR imagery can be geocoded with an accuracy of 30-60 m using ground control points. With regard to directional wave spectra, AS AR's wave mode to detect waves in the open oceans will probably be of limited use in shallow seas, where wavelengths are much shorter.

PHARUS is a polarimetric phased array C-band Synthetic Aperture Radar (SAR), designed and built for airborne use. Advanced SAR (ASAR) data in image and alternating polarization mode have been simulated with PHARUS to demonstrate the use of Envisat for a number of typical SAR applications that are not possible with ERS. These applications concern landuse classification with several combinations of polarimetric channels and terrain height estimation from a same side stereo pair. In addition, SAR image geocoding has been investigated as well. PHARUS data collected over the Black Forest in Germany and over a cultivated area in The Netherlands were degenerated in the processing to ASAR resolution with two looks in alternating polarization mode and four looks in image mode. Results show that ASAR dual-channel data from alternating polarization mode can better separate vegetation types than the single-channel data of the ERS satellites. Moreover, with ASAR resolution, terrain relief can be estimated with an accuracy of about 50 m, whereas SAR imagery can be geocoded with an accuracy of 30-60 m using ground control points.

The PHARUS system is a fully polarimetric C-band SAR with an active phased array anrenna. PHARUS is an experimental system. It is meant for remote sensing research in many application areas, both civil and military, maritime and on land. The system performed its first test flight in 1995. In the following years a large amount of data was acquired. In the standard mode the resolution in both range and azimuth is about 3 m. the capability of PHARUS to obtain fully polarimetric images with a high resolution is extensively used. More advanced SAR applications require both high resolution and the flexibility that is offered only by phased array SAR systems The PHARUS system, since it hs an active array, which is fully programmable in its operaring modes allows investigations of these applications. Advanced SAR modes that are being investigated are: (sliding) spotlight SAR and interferometric SAR, both along track (Moving Target Indication) and across track (Repeat Pass Interferometry). The flight experiments are carried out in co-operation with the National Aerospace Laboratory. In this paper the status of the different new processing techniques is reported. These new techniques include doubling of the range resolution, increasing the azimuth resolution through spotlight processing, Displaced Phase Center Antenna MTI and Repeat Pass Interferometry. Also the use of PHARUS as a test bed for the ASAR modes of the ENVISAT satellite is shown.

We have studied an optimal target detection procedure for polarimetric SAR data by using PHARUS data collected during the MIMEX campaign. The detection method is especially suitable when no a priory knowledge of the target is available. We have found that polarimetric whitening filtering preceding an order-statistics CFAR detector gives the best results for target detection. We also found that polarimetric information can support target recognition and classification. For terrain interpretation polarimetry is quite indispensable. For target detection spatial resolutions in the order of 1-5 meter are sufficient, while for recognition and classification a resolution less than I meter is required.