helios icon indicating copy to clipboard operation
helios copied to clipboard

Coordinate System Definitions in HELIOS++

Open lwiniwar opened this issue 4 years ago • 0 comments

Currently, HELIOS++ assumes a local planar coordinate system with all inputs in the same CRS (especially when working with GeoTiffs). This leads to the following limitations:

  • Earth curvature is ignored
  • No proper modelling of satellites (IceSAT, GEDI/ISS) possible
  • Possible issues with large-scale (100km+) surveys
  • Unit clash (e.g. if TIFF is in EPSG:4326)

A proper solution could be transformation of everything into ECEF (Earth Centered Earth Fixed) - i.e. a cartesian 3D coordinate system, like it is being used internally in RiProcess and RiScan Pro. Problems:

  • Transformation of "linear" segments (e.g. legs) from a Projected CRS result in 3D curves in ECEF - here, one would need to define a sampling interval/linearization step
  • Potentially big coordinates need to be treated somehow (e.g. BIGINT)
  • For the "simple" cases, we transform back and forth -> runtime

Also see the coordinate system definitions in OPALS: https://opals.geo.tuwien.ac.at/html/stable/ModuleDirectGeoref.html#dirgeoref_world_system

lwiniwar avatar Jan 22 '21 08:01 lwiniwar