The output of geocoding has a patchy structure on the pixel scale, what happened?

[rese_support]

The small scale structure depends on the type of interpolation you choose in the final processor. For radiometric reasons, nearest neighbour approaches are recommended for resampling (any interpolation results in a loss of information). Nearest neighbour bears the disadvantage of a patchy structure on the pixel-scale. Some compromise solutions are implemented in PARGE. The 'nearest bilinear' method is recommended since it combines nearest neighbour and interpolation methods. If the focus is on highest spatial accuracy, the bilinear interpolation type has to be chosen.