changing between Geoids and Ellipsoids with lasheight

1,884 views
Skip to first unread message

Martin Isenburg

unread,
Feb 13, 2012, 11:54:34 AM2/13/12
to LAStools - efficient tools for LIDAR processing
Hello,

a while ago I had a French university researcher contact me on using
LAStools to convert from ellipsoid heights to a Geoid model. The Geoid
he used was computed by the IGN (Institut Géographique National) and
was provided as ASCII points: x and y were Longitude and Latitude, z
was the elevation of the ellipsoid in respect to that of the geoid in
meter (in France, the Geoid is about 40 - 50 m above the ellipsoid).
Here is the workflow he used. He first cut the Geoid file "geoid.txt"
around area of interest and converted it to LAS

txt2las -i geoid.txt -iparse xyz -clip xmin ymin xmax ymax -
change_classification_from_to 0 31 -set_scale 1e-9 1e-9 0.001 -o
geoid.las

He choose class 31 for the points from the Geoid because he had no
points with class 31 in his LAS files and set the scale to equal that
of his LAS files. Then for each data file (data01.las
data02.las ....), he did:

lasheight -i dataXX.las geoid.las -merged -rescale -class 31 -o
dataXX_with_geoid.laz -replace_z

Finally, he dropped out the Geoid points:

las2las -i dataXX_with_geoid.laz -o dataXX_final.laz -drop_class 31

For the first test he had differences of a few millimeters between the
obtained elevations values and the ones obtained with another software
(that took hours instead of seconds for the computation). This
difference was due to the different interpolation method (TIN versus
bilinear). He then densified the Geoid model by adding points computed
by bilinear interpolation between the given points. By dividing the
mesh size by 10 both in longitude and latitude, he could lower the
differences to less than 1 mm ...

If there is big demand for doing such "explicit grid" based Geoid
transforms, I could probably add an option to lasheight that would
allo doing this in one step.

Cheers,

Martin @lastools

Martin Isenburg

unread,
Feb 15, 2012, 8:38:33 AM2/15/12
to LAStools - efficient tools for LIDAR processing
Hello,

after describing how to convert between Geoids and Ellipsoids with
lasheigh I got this message: "I think I am the researcher you are
talking about. I can confirm you, after having used this method quite
intensively, that It gives correct results and is very time efficient.
I wrote my own visual basic macro in order to do the different steps
but I think it will be nice and useful for myself and other people if
this could be added to lasheight. I suppose that many people use a
geoid model to convert ellipsoid heights to their local elevation
coordinates."

http://lastools.org/download/lasheight_README.txt
http://lastools.org/download/lasheight.exe

So I added this functionality in the latest release (120214).

lasheight -i in.las -ground_points geoid.txt -replace_z -o out.las

You can also convert an entire full folder of LAS / LAZ / BIN / TXT
files

lasheight -i *.las -ground_points geoid.xyz -replace_z -odix _geoid

and you can do this with multiple cores (but better use LAZ to fight
the I/O bottleneck).

lasheight -i *.laz -ground_points geoid.las -replace_z -odix _geoid -
olaz -cores 4

I asked said researcher to try this out and he wrote: "I have tried
this new option and it does work very well for me."

Cheers,

Martin @lastools

Khamarrul Azahari Razak

unread,
Feb 15, 2012, 9:16:18 AM2/15/12
to last...@googlegroups.com

It is indeed very useful. Thanks Martin for this update.

Khamarrul

Hello,

http://lastools.org/download/lasheight_README.txt
http://lastools.org/download/lasheight.exe

Cheers,

Martin @lastools

--
Download LAStools at
http://lastools.org/
Visit the LAStools group at
http://groups.google.com/group/lastools/
Be social with LAStools at
http://www.facebook.com/LAStools
http://www.twitter.com/LAStools

mana...@gmail.com

unread,
Mar 28, 2017, 6:48:50 AM3/28/17
to LAStools - efficient tools for LiDAR processing, ra...@itc.nl

Hello,


First of all apologize if the doubt that I raise is too basic or has already been raised in the forum. But looking for topics in the forum I did not find the concrete solution to my problem.

I was trying to convert a LIDAR point cloud from an ellipsoid height model to an orthometric accurate height model.  The geoid model in Spain is EGM08-REDNAP. Calculated by de IGN (Instituto Geográfico Nacional) adapting the world gravity model EGM2008 to the Vertical Reference System in Spain materialized by the High Precision Levelling Network (REDNAP).


You can download the geoid model from IGN website in an .asc file that contains a regular grid (1’x1’) with the geoid undulation values, described by rows begining at northwest corner. The file has a heading  and it is simlilar to a Digital Terrain Model.


I saw in the README file and in some forum entries
, that the lasheight tool can be used to convert ellipsoid heights to orthometric heights.


 lasheight -i in.las -ground_points geoid.txt -replace_z  -o out.las 


But in my case it doesn’t work because, I suppose, the geoid model file doesn’t have the information it has to contain for the tool to work.  

Is there a way to convert .las ellipsoid height point cloud into an orthometric point cloud using this model file?

 

Thank you very much.

Martin Isenburg

unread,
Mar 28, 2017, 9:58:53 AM3/28/17
to LAStools - efficient command line tools for LIDAR processing
Hello,

If you have the GEOID in ASC raster format *and* your GEOID and your LiDAR have the same horizontal projection then your command line should work.

lasheight -i in.laz -ground_points geoid.asc -replace_z  -o out.laz


However, it would be more efficient to convert the GEOID to the LAZ format and sort it into a space-filling curve that is then spatially indexed as follows:


lassort -i geoid.asc -o geoid.laz
lasindex -i geoid.laz

and then run

lasheight -i in.laz -ground_points geoid.laz -replace_z  -o out.laz

not sure if the two have the same horizontal projection? Just have a look at them in the LAStools GUI:

lasview -i geoid.laz -i in.laz -gui

you should see two bounding boxes. One gigantic (the GEOID) and one tiny (the LiDAR) and the latter should be somewhere within the first. 

Regards,

Martin @rapidlasso


Be social with LAStools at
Reply all
Reply to author
Forward
0 new messages