I'm working on Windows 10. In input, I have a .LAS "GROUND_only" LIDAR point cloud (4 millions points). We're interested in geomorphological microstructures called raised field (ground elevation due to cultivated areas). See picture below:

Here is a screenshot representing this cloud file - broad view

We have generated a DEM raster file (planimetric resolution 20 cms from this cloud point). Then, we have generated contour lines (10 cms equidistance). I paste below a screenshot representing these contour lines. The difference between the top of the raised fields and the bottom is often 20/40 cms:

Here is a screenshot representing this cloud file - close viewe

  1. We would like to classify our input ground point cloud in order to separate the points constituting those "raised field" and other ground points "NON ground raised field" points

Once the classification is done, we would like to be able to export the result toward a new .las or shp (ideally with useful attribute information like class_id, z value, etc.).

We would like generate a Multipoint layer only for those "raised field objects".

All the points constituting an object "raised field" must be stored into the same row (entity).

Could you tell me how to that?

  • I have only listed the soft likely to be useful : lastools,grass,etc but which algorithm? which method?....Those raised fields have a minimum height of 30 cms (up to 60 cms). In order to identify the raised field class, we thought that slope (angle) also would be an interesting criterion for discriminate..In a first time, our goal is to isolate the raised fields element. In a second time, we wd like generate a concave hull from MULTIPOINTS layer in order to have the spatial limit of those raised field (we know do this step without problem with postgis with st_concave_hull ). Thanks.
    – user35117
    Mar 22, 2017 at 6:10
  • Have you tried smoothing the DEM (for example, gis.stackexchange.com/questions/9431/…), and then, subtract the smoothed DEM from the original DEM? (something like: gis.stackexchange.com/questions/2555/…)? Mar 28, 2017 at 20:20


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.