Minimize Processing Overview

The main aim of Level-1 synergy processing is the estimation of de-registration between the measurements from OLCI and those from SLSTR. There are several steps which outline the concept of this processing.

  • Two channels with very similar wavelengths are selected, one from OLCI and one from SLSTR (Oq and Su). The channels are user selectable with the default channels being Oq=Oa17 and Su=S3.
  • Ortho-geolocation information from Oq is regarded as the reference.
  • Ground control point pixels in Su are expressed as a function of the ortho-geolocation from Oq and a subsequent inverse ortho-geolocation of Su, establishing a link between Oq and Su pixels.
  • A small block of pixels around this ground control point in Su is projected on the Oq grid (this grid will also serve as the final common grid).
  • The block of interpolated Su pixels on the Oq grid is compared to a block of Oq pixels by calculating the cross-correlation between the two blocks (windows).
  • The block of Oq pixels (context window) is moved according to shift vectors. At each of the shifts, the cross-correlation with the Su window (search window) is calculated. The elements of the shift vectors at which a maximum cross-correlation is reached determine the pixel de-registration between Su and Oq.
  • A sub-pixel de-registration is sought. At each ground control point, the cross-correlation is interpolated on a finer grid and its maximum is found. This process is repeated a number of times (or a convergence criterion is applied), yielding the sub-pixel de-registration at each ground control point.
  • De-registration information for all pixels is obtained by applying a 2D polynomial to the obtained de-registrations at the ground control points (this is done separately for different regions in the image).

Besides the core processing described, a number of accompanying steps are necessary and a general overview of all processing steps is shown in figure below.

Overview of Level-1 Synergy processing steps

Note: The core of the whole Synergy Level-1 processing, the estimation of de-registration between two reference channels, is repeated for each OLCI camera as well as for each ground control point.