Monitoring harbour activity from (high-resolution) SAR imagery

Functie

Background
Satellite-based Synthetic Aperture Radar (SAR) holds clear advantages over more conventional optical and multispectral sensor observations. In particular, the advantage of cloud penetrating signals allows for consistent monitoring, object- and change detection. SAR day- and night images allow for continuous and non-intrusive monitoring of harbour areas, providing a comprehensive view of activities over time. This enables managing (illegal) dock activity, identifying marine resources (e.g., fishing boats), optimizing cargo logistics and thus assisting in maritime operations. However, complexity of SAR-signals, associated with variable configuration settings between acquisitions, prohibits easy quantitative interpretation and qualitative detections for monitoring applications. By elucidating on the effect of sensor configurations and signal processing we aim to support operational supervision and situational awareness to enhance safety and security in harbours.

Assignment
The objective of this assignment is to assess pre-processing steps, that are commonly taken for SAR images, on the detection capability of monitoring harbour activity. Harbour activity of interest can include, but are not limited to, ship movement, container transport and oil levels in tanks.

We envision the following actions during the thesis period:

• Literature study of the effect of different SAR-processing concepts, in particular:
• • speckle filtering
  • object detection
  • object classification
• Compare the quality of different types of SAR observations, including RadarSAT-2, TerraSAR-X and new generation constellations like Iceye, Capella or Umbra
• Development of proof-of-concept(s) that include the processing and analysis of SAR-images for monitoring harbour activities an and in particular the effect of different pre-processing steps on the detection capability.
• Analysis of temporal and spatial changes occurring in the port of Rotterdam.

Result
Different preferred topics are described above. In agreement with the student’s interest these can be further refined. Regardless of the chosen topics the following outcome of the internship is expected:

• A report on the selected points of the assignment section above.
• A proof-of-concept processing chain.

Duration
6-8 months

Profiel

Requested level

• Masters student in one of the following fields:
  • Physical Geography / Geology
  • Earth Observations / Remote Sensing
  • Data science / Artificial intelligence
• Proficiency in Python programming
• Interest in image processing and remote sensing, preferably SAR.
• Assertiveness and a can-do attitude to make it a successful project

Arbeidsvoorwaarden

What we offer

• A challenging internship project in a high-tech result orientated work environment, either at NLR in Amsterdam or Sensar in Rijswijk
• Multidisciplinary and international team of supervisors and colleagues, both at NLR and Sensar
• Weekly supervision and availability of the technical staff for support of both NLR and Sensar
• An internship allowance

Informatie

About NLR
Royal NLR operates as an objective and independent research centre, working with its partners towards a better world tomorrow. As part of that, NLR offers innovative solutions and technical expertise, creating a strong competitive position for the commercial sector.

About Sensar
The current boom in the big data and satellite industries has the potential to revolutionize the civil engineering monitoring industry and at Sensar we work exactly on this interface. We believe in grounded decisions, no more surprises. We want to enable experts in the civil engineering industry, i.e., municipalities, waterboards, construction companies, to use satellite data in analyzing and managing their processes by making this data easily accessible, easy to use and attractively priced. In our cloud-based platform we use the InSAR technique to produce monitoring products for civil engineers, allowing them to always have access to the most recent information on their assets.

Solliciteren

Interested?
Get in touch by sending your motivation and resume to Job de Vries, job.de.vries@nlr.nl or Hannah Min Hannah.min@nlr.nl.