Day 4

Detailed paper information

Back to list

Paper title GLANCE: A Multi-sensor Geolocation and Co-registration Verification Toolbox
  1. Marcos Bento HE Space Operations @ ESA/ESTEC Speaker
  2. Sergio Bras HESpace Operations for ESA Speaker
  3. Alberto Graziani European Space Agency - ESA-ESTEC
  4. Ines Fuente ESA/ESTEC, Noordwijk, Netherlands
  5. Fermin Alvarez Lopez ESTEC/ESA
  6. Berthyl Duesmann European Space Agency/ESTEC
Form of presentation Poster
  • Open Earth Forum
    • C5.03 Open Source, data science and toolboxes in EO: Current status & evolution
Abstract text The evaluation of multi-sensor image geolocation and co-registration is a fundamental step of any mission commissioning and in-orbit validation (IOV) phases, allowing to detect possible misalignments between instruments and platform introduced by the vibrations that happen during launch. At these stages, information about existing misalignments is crucial to support sensor calibration activities. The image evaluation considers a quantitative analysis of remote sensing data, typically L1b products acquired over different earth locations, in relation to a reference image composed of reference geographical features (Geolocation) or to data produced by another sensor aboard (Co-registration).

This work addresses the design, development, verification and operation of the set of tools provided by Geolocation and Co-registration (GLANCE) toolbox, which will be used during the in-orbit verification activities of the MetOp-SG mission. The MetOp-SG is the space segment of the EPS-SG (EUMETSAT Polar System Second Generation) and it is comprised of two spacecraft (MetOp-SG A and B). During the MetOp-SG IOV activities the geolocation and co-registration checks for which GLANCE will be used focuses on evaluating the L1b data provided by Sentinel-5, 3MI, METimage, MWS and IASI-NG (MetOp-SG A), and MWI and ICI (MetOp-SG B). GLANCE enables several combinations of multiple sensor images for purposes of geolocation: 3MI, MWS, MWI, ICI; and co-registration: 3MI vs Sentinel-5, and METimage vs Sentinel-5. Although originally designed for MetOp-SG, GLANCE leverages on a generic component-based architecture to provide a comprehensive toolbox regarding image processing functionality (i.e. image convolution, edge detection, thresholding) which can easily be extended to support other sensors and/or missions. GLANCE integrates several capabilities available in open source packages (e.g. OpenCV, GDAL) with specifically designed functionality, such as the generation of reference images based on geographical characterization data.
Considering that each sensor image has specific processing requirements, GLANCE enables the user to compose multiple processing steps into a customizable processing chain, and in order to perform batch processing of images, GLANCE applies autonomously the transformations specified by a processing chain, evaluating the existence of misalignments without human intervention. This automatic processing capabilities are necessary, considering the reduced time extent available for IOV activities. Moreover, GLANCE design and development has taken runtime performance to process the expected IOV images into consideration and takes advantage of parallelism whenever supported by the hardware.
After describing the context of operations during the IOV phase, including the interaction with other ground segment components, the GLANCE toolbox architecture and design will be introduced along with the description of the processing steps performed while evaluating the images of the multiple sensors. The catalogue of processing capabilities and algorithms will be presented together with preliminary results.