3

I've been struggling with this for the past few days. I want to create a DEM of a flood control area, which is covered in low/intermediate-height vegetation (pioneers and reed mostly). The point cloud data with x, y and z comes from airborne LiDAR and I've been able to load it into QGIS by conversion and translation and managed to create a DSM.

From false-colored images, I know the locations where the vegetation is present. Is there a way to get an actual DEM of the area by removing the low/intermediate-vegetation? My aim is to obtain a DEM to analyze the flooding area, so an increase in the accuracy of the elevation data would be very useful.

Part of the headerless .txt file LIDAR; so x coord, y coord and z coord / elevation;

147675.500 210500.500 10.15
etc Ln 1805277, Col 1 (Notepad)

This is how the DTM looks after I've loaded the text file via Raster>Conversion>Translation and used a IDW-interpolation:

enter image description here

This is de CIR image from the area, showing the vegetation cover:

enter image description here

I thought when using LASTools I could extract the bare ground, so first I've used txt2las to convert the .txt file to a .las file. When running the lasinfo on this newly created .las file, this is what I get:

reporting all LAS header entries: 
file signature: 'LASF' 
file source ID: 0 
global_encoding: 0 
project ID GUID data 1-4: 00000000-0000-0000-0000-000000000000 
version major.minor: 1.2 
system identifier: 'LAStools (c) by rapidlasso GmbH' 
generating software: 'txt2las (version 200101)' 
file creation day/year: 70/2021 
header size: 227 
offset to point data: 321 
number var. length records: 1 
point data format: 0 
point data record length: 20 
number of point records: 1805276 
number of points by return: 1805276 0 0 0 0 
scale factor x y z: 10 10 10 
offset x y z: 0 0 0 
min x y z: 147680 210500 -10 
max x y z: 150000 213000 20 
variable length header record 1 of 1: 
reserved 0 
user ID 'LASF_Projection' 
record ID 34735 
length after header 40 
description 'by LAStools of rapidlasso GmbH' 
GeoKeyDirectoryTag version 1.1.0 number of keys 4 
key 1024 tiff_tag_location 0 count 1 value_offset 1 - GTModelTypeGeoKey: ModelTypeProjected 
key 3072 tiff_tag_location 0 count 1 value_offset 31370 - ProjectedCSTypeGeoKey: Belge 1972 / Belgian Lambert 72 
key 3076 tiff_tag_location 0 count 1 value_offset 9001 - ProjLinearUnitsGeoKey: Linear_Meter 
key 4099 tiff_tag_location 0 count 1 value_offset 9001 - VerticalUnitsGeoKey: Linear_Meter 
reporting minimum and maximum for all LAS point record entries ... 
X 14768 15000 
Y 21050 21300 
Z -1 2 
intensity 0 0 
return_number 1 1 
number_of_returns 1 1 
edge_of_flight_line 0 0 
scan_direction_flag 0 0 
classification 0 0 
scan_angle_rank 0 0 
user_data 0 0 
point_source_ID 0 0 
number of first returns: 1805276 
number of intermediate returns: 0 
number of last returns: 1805276 
number of single returns: 1805276 
overview over number of returns of given pulse: 1805276 0 0 0 0 0 0 
histogram of classification of points: 
1805276 never classified (0) 

So none of the points is classified as anything.

I use QGIS 3.16 and the LiDAR data I received is in headerless .txt format. I've tried LAStools and converted the LiDAR .txt into .las, but the las2dem is not giving an output file and says:

WARNING: unlicensed. over 1.5 million points. inserting black diagonal.
ERROR: cannot find triangle in first 10000 points.

As you can probably tell, I'm very much a beginner in QGIS.

6
  • 1
    Interesting. Typically, the DSM is derived from the first return or from Class Code = 5 (high vegetation). Concomitantly, the DTM (aka DEM) is derived from the last return or Class Code = 2 (Bare earth). Can you confirm that your source data has those latter values? – Stu Smith Mar 11 at 15:44
  • Thank you for your comment! I tried to get info of any classifications via the the lasinfo tool (after I used the text2las on the .txt file LiDAR data). However, since the .txt file only has the three columns of x,y,z coords, the following conversion to .las doesn't seem to have any classifications (yet). Can the classification be added after? I hope my answer makes a bit of sense, otherwise please let me know. – Marel_S Mar 11 at 20:28
  • I'm a little confused. What return value or class code did you use to create the DSM? – Stu Smith Mar 11 at 20:31
  • I suggest that you edit your original question to show exactly, step-by-step, how you successfully created the DSM. Include a screenshot of the DSM. Then edit to show the steps you took to attempt the DTM (DEM) creation. Include the error statements. – Stu Smith Mar 11 at 21:09
  • I've added some info, hope this helps. Thanks again! – Marel_S Mar 12 at 13:13
2

You can extract bare ground with lastools.

On the newly created las file, you have to run lasground => it will find the ground points. You might need to play with the parameters of lasground (see here) until you are happy with the result (can always visualise them with lasview).

Once you are happy with the ground points extraction, run blast2dem on your classified las point cloud with -keep_class 2 - it will build a DEM based only on the points identified as ground.

Good luck and let us know if this helps you.

2
  • Thank you for your comment. I've tried this before, but somehow LAStools does not let me > Save to File or > Save to Temporary file. LAStools is in C:\LAStools – Marel_S Mar 16 at 8:15
  • Are you using the GUI? It might be easier to use the command line interface. First add "C:\LAStools\bin" to your PATH, then you can run e.g. "lasground -i input.las -o output.las" directly from the working directory where the files are located. – Marcel Mar 16 at 10:16
0

Ok. As far as I can tell you need to classify your pointcloud first.

For ground points classification you can use FUSION lidar GroundFilter tool which will classify your point clouds to 2 categories ground points and all the rest. From this you can create a DTM. I have this short tutorial on how to make DTM from classified pointcloud: https://arheologija.neocities.org/Lidar_tutorial.html#4.

Whole tutorial is based on Slovenian lidar data, but its the same workflow for any classified pointcloud. If you have any other questions, ask away :)

Your Answer

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

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