Share Email Print

Proceedings Paper

Rugged: an operational, open-source solution for Sentinel-2 mapping
Author(s): Luc Maisonobe; Jean Seyral; Guylaine Prat; Jonathan Guinet; Aude Espesset
Format Member Price Non-Member Price
PDF $17.00 $21.00

Paper Abstract

When you map the entire Earth every 5 days with the aim of generating high-quality time series over land, there is no room for geometrical error: the algorithms have to be stable, reliable, and precise. Rugged, a new open-source library for pixel geolocation, is at the geometrical heart of the operational processing for Sentinel-2. Rugged performs sensor-to-terrain mapping taking into account ground Digital Elevation Models, Earth rotation with all its small irregularities, on-board sensor pixel individual lines-of-sight, spacecraft motion and attitude, and all significant physical effects. It provides direct and inverse location, i.e. it allows the accurate computation of which ground point is viewed from a specific pixel in a spacecraft instrument, and conversely which pixel will view a specified ground point. Direct and inverse location can be used to perform full ortho-rectification of images and correlation between sensors observing the same area. Implemented as an add-on for Orekit (Orbits Extrapolation KIT; a low-level space dynamics library), Rugged also offers the possibility of simulating satellite motion and attitude auxiliary data using Orekit's full orbit propagation capability. This is a considerable advantage for test data generation and mission simulation activities. Together with the Orfeo ToolBox (OTB) image processing library, Rugged provides the algorithmic core of Sentinel-2 Instrument Processing Facilities. The S2 complex viewing model - with 12 staggered push-broom detectors and 13 spectral bands - is built using Rugged objects, enabling the computation of rectification grids for mapping between cartographic and focal plane coordinates. These grids are passed to the OTB library for further image resampling, thus completing the ortho-rectification chain. Sentinel-2 stringent operational requirements to process several terabytes of data per week represented a tough challenge, though one that was well met by Rugged in terms of the robustness and performance of the library.

Paper Details

Date Published: 12 October 2015
PDF: 10 pages
Proc. SPIE 9639, Sensors, Systems, and Next-Generation Satellites XIX, 963924 (12 October 2015); doi: 10.1117/12.2194344
Show Author Affiliations
Luc Maisonobe, CS Systèmes d'information (France)
Jean Seyral, CS Systèmes d'Information (France)
Guylaine Prat, CS Systèmes d'Information (France)
Jonathan Guinet, CS Systèmes d'Information (France)
Aude Espesset, CS Systèmes d'Information (France)

Published in SPIE Proceedings Vol. 9639:
Sensors, Systems, and Next-Generation Satellites XIX
Roland Meynart; Steven P. Neeck; Haruhisa Shimoda, Editor(s)

© SPIE. Terms of Use
Back to Top
Sign in to read the full article
Create a free SPIE account to get access to
premium articles and original research
Forgot your username?