I have some multi spectral imagery (raster data) from a UAV of vineyards in Victoria, as well as corresponding 3D point cloud (also raster) of the vineyard topography, based off ultrasonic sensing, which picks up the height difference between the vines and the ground between the rows.

The article linked below discusses a method for extracting the vine pixels using the height differentials expressed in the point cloud.


Specifically they divide the resulting imagery up into a grid consisting of 10m x 10m squares and calculate a reference height for each square (using to lowest 10% of pixels in the point data). Pixels exceeding the reference height reflect the vine pixels and are extracted.

I'm looking for some guidance as to how I might do this using QGIS. I'd like to generate a binary mask that could be used to say, extract NDVI data for the vines only.

I'm fairly well versed with QIS and how to drive it, but this has me stumped.

  • I don't have a full answer for you, but you may want to look into SAGA GIS, which is also open source. It is accessible through the QGIS toolbox with a little configuration, but is sometimes easier to use directly. A Google search for "NDVI SAGA GIS" gives some results that may be promising for you. – Nate Wanner Mar 16 '17 at 1:55
  • I found this tutorial helpful to understand point cloud processing. – Nate Wanner Mar 16 '17 at 2:05

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.