6

I´m trying to extract DEM and DTM from a LIDAR file. Currently, I rasterize the image but don´t know which algorithm I need to apply to obtain the DEM and DTM information in different models to be able to compare both of them.

Improve this question
2
  • What is your input file type ? typically if you use the first return you will have the DTM already.
    – radouxju
    Mar 16, 2015 at 15:52
  • 3
    eCognition is not remotely designed for this type of analysis. Whereas methodology such as watershed segmentation have been used to identify ground returns this was a specialized algorithm and segmentation on a single image is not an appropriate approach. It is time for you to dig into the literature to understand how lidar filtering works and settle on an appropriate methodology.
    – Jeffrey Evans
    Jul 17, 2015 at 17:41

5 Answers 5

Reset to default
1

I've recently completed this same analysis, just in ArcMap 10.1-10.3; you won't need eCognition, although it can create segmented object-based DSM/DTM.

  1. add the .LAS file to an LASD dataset in arc catalog
  2. convert to raster using the LAS toolset mentioned by Cameron above
  3. using the raster calculator tool (LAS elevation) - (DEM), which should be your DSM/DTM.

Next you can reclassiffy these accordingly, including masking out other unwanted noise (e.g. set a threshold for the heights you're looking for)

Improve this answer
0

I came across a toolset that I have used for DEM and DTM extractions called the LAS Toolset. I use it in ArcGIS, however I believe that it also can be used independently.

The download for the toolset is found at : http://www.cs.unc.edu/~isenburg/lastools/

Through this you can use filters to filter out all of the LAS points that are not ground/water for the DEM. The DSM will have less filters placed according to what you are looking to capture. Once this is performed you should be able to save it in a different number of files. Ideally you can use file converters to bounce around to what you need, however keep in mind that the more direct, the less chance of file corruption.

This is what I have found from my personal experiences with LiDAR data extraction. I hope that this helps!

Improve this answer
1
  • 1
    A DSM or DTM would represent an unfiltered, interpolation of the point cloud were the DEM is the filtered interpolated ground or "bare earth" surface.
    – Jeffrey Evans
    Jul 17, 2015 at 17:43
0

Having done this with both airborne and terrestrial LiDAR, I would suggest using Whitebox.

It has a fairly robust set of tools specifically for handling LiDAR data and it's quite fast at handling LAS files and can remove non-ground objects. Each tool has a pretty good description of how to use it and what options to select.

It is also cross-platform (if you aren't on Windows).

Improve this answer
0

There is tutorial reagarding this. Link to convert point cloud data to DEM, DTM, nDSM

eCognition Developer Tutorial 8 — Working with Point Clouds in eCognition

I hope this helps.

Improve this answer
0

the current version of eCognition provides Point Cloud Algorithms to facilitates extract information from point cloud data.

The 'Rasterize point cloud' (Former name: LiDAR File Converter.) algorithm lets you rasterize point clouds (.las or .laz files) into a 2D raster layer by using point fields such as elevation and number of returns. This information will be added to the project as additional raster layers. Multiple rasterizations can be processed for the same point cloud input.

The Quick Terrain Modeler is a powerful package to process point cloud data and derive DTM/nDSM. The derived products can be analyzed within eCognition if you are preferring to work with image objects.

Improve this answer

Your Answer

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

Not the answer you're looking for? Browse other questions tagged or ask your own question.