Las2dem: Setting appropriate "step size" & "thin with grid" values according to the point density of the initial LIDAR data
534 views
Skip to first unread message
Kim
Jul 6, 2018, 9:58:27 AM7/6/18
to LAStools - efficient tools for LiDAR processing
Dear Lastools community,
My study consists in mapping roofs of small buildings in a really dense urban area.
Unfortunately, the Las files provided have a low point density:
number of first returns: 1975993
number of intermediate returns: 4686
number of last returns: 1975459
number of single returns: 1858098
covered area in square units/kilounits: 1236116/1.24
point density: all returns 1.70 last only 1.60 (per square units)
spacing: all returns 0.77 last only 0.79 (in units)
In las2dem, I would like to set the most precise step size (possible with the point density of my las files) in order to produce my DEM and DTM.
I thought this combination could be good:
step size of 0.3 with a "thin with grid" 0.1
(my tile size is 1000)
I don't know if I can go even smaller than 0.3 as a step size since my point density is only 1.60/m2 (last only).
Could please someone explain me how to set a precise step size according to the point density?
Thank you!
Martin Isenburg
Jul 7, 2018, 6:45:42 AM7/7/18
to LAStools - efficient command line tools for LIDAR processing
Hello Kim.
from your lasinfo report output I read that you have an average of 1.6 laser pulses / last returns per square unit (assuming square meter) which translates into an
average spacing of 0.79 meters between the laser pulses / last returns. There is certainly no need to thin this already low density data on-the-fly onto a 10 centimeter grid as the option '-thin with grid 0.1' would do. I don't think this will do anything but cost computation and memory resources.
I think with your pulse spacing you can get a decent 1 meter DTM / DSM using a step size of '-step 1.0' and maybe there is a bit more information to be gained by setting the step size to 0.8 or even 0.5 but I doubt that 0.3 is a useful raster resolution for this low density. Note that in generate 0.3 is a bad choice in resolution for a 1000 by 1000 tile because you cannot divide 1000 by 0.3 without a remainder. Seen from that angle 0.2, 0.25, 0.4, 0.5, 0.8, or 1.0 should be your only considered choices.
The average density may not even reflect the true density as there could be one part of the tile covered by two flight lines and another covered only by one (or similar). To get a more detailed look at your densities you can create visualizations with lasgrid or las2dem
http://lastools.org/download/lasgrid_README.txt
E:\LAStools\bin>lasgrid -i ..\data\france.laz -last_only -density -step 0.5 -false -set_min_max 0 15 -odix _d_0_15 -opng
http://lastools.org/download/las2dem_README.txt
E:\LAStools\bin>las2dem -i ..\data\france.laz -last_only -edge_shortest -step 0.5 -false -set_min_max 0.1 0.4 -odix _shortest_10cm_40cm -opng
E:\LAStools\bin>las2dem -i ..\data\france.laz -last_only -edge_longest -step 0.5 -false -set_min_max 0.3 0.8 -odix _longest_30cm_80cm -opng
An exact histogram of edge lengths of the last return TIN will approximate the true distribution of inter-pulse distances as good as possible with las2tin.
E:\LAStools\bin>las2tin -i ..\data\france.laz -last_only -histo_only edge_length 0.1
edge length histogram histogram with bin size 0.100000
bin [0.0,0.1) has 10188
bin [0.1,0.2) has 27961
bin [0.2,0.3) has 45157
bin [0.3,0.4) has 58498
bin [0.4,0.5) has 86948
bin [0.5,0.6) has 29746
bin [0.6,0.7) has 12542
bin [0.7,0.8) has 4455
bin [0.8,0.9) has 1245
bin [0.9,1.0) has 310
bin [1.0,1.1) has 154
bin [1.1,1.2) has 59
bin [1.2,1.3) has 48
bin [1.3,1.4) has 48
bin [1.4,1.5) has 42
bin [1.5,1.6) has 32
bin [1.6,1.7) has 13
bin [1.7,1.8) has 23
bin [1.8,1.9) has 25
bin [1.9,2.0) has 17
bin [2.0,2.1) has 15
bin [2.1,2.2) has 18
bin [2.2,2.3) has 12
bin [2.3,2.4) has 8
bin [2.4,2.5) has 13
bin [2.5,2.6) has 10
bin [2.6,2.7) has 12
bin [2.7,2.8) has 10
bin [2.8,2.9) has 12
bin [2.9,3.0) has 9
bin [3.0,3.1) has 8
bin [3.1,3.2) has 5
bin [3.2,3.3) has 11
bin [3.3,3.4) has 3
bin [3.4,3.5) has 3
bin [3.5,3.6) has 7
bin [3.6,3.7) has 5
bin [3.7,3.8) has 7
bin [3.8,3.9) has 9
bin [3.9,4.0) has 4
bin [4.0,4.1) has 5
bin [4.1,4.2) has 5
bin [4.2,4.3) has 4
bin [4.3,4.4) has 8
bin [4.4,4.5) has 5
bin [4.5,4.6) has 4
bin [4.6,4.7) has 5
bin [4.7,4.8) has 4
bin [4.8,4.9) has 4
bin [4.9,5.0) has 3
bin [5.0,5.1) has 6
bin [5.1,5.2) has 4
bin [5.2,5.3) has 4
bin [5.3,5.4) has 3
bin [5.4,5.5) has 3
bin [5.5,5.6) has 3
bin [5.6,5.7) has 3
bin [5.9,6.0) has 4
bin [6.0,6.1) has 5
bin [6.1,6.2) has 1
bin [6.2,6.3) has 2
bin [6.3,6.4) has 4
bin [6.4,6.5) has 2
bin [6.5,6.6) has 5
bin [6.7,6.8) has 1
bin [6.8,6.9) has 5
bin [6.9,7.0) has 2
bin [7.0,7.1) has 3
bin [7.1,7.2) has 4
bin [7.2,7.3) has 3
bin [7.3,7.4) has 4
bin [7.4,7.5) has 1
bin [7.5,7.6) has 3
bin [7.6,7.7) has 3
bin [7.7,7.8) has 3
bin [7.9,8.0) has 3
bin [8.0,8.1) has 1
bin [8.1,8.2) has 2
bin [8.2,8.3) has 1
bin [8.3,8.4) has 2
bin [8.4,8.5) has 1
bin [8.6,8.7) has 2
bin [8.7,8.8) has 2
bin [8.9,9.0) has 2
bin [9.1,9.2) has 3
bin [9.2,9.3) has 1
bin [9.5,9.6) has 1
bin [9.6,9.7) has 3
bin [9.7,9.8) has 2
bin [9.8,9.9) has 1
bin [10.0,10.1) has 1
bin [10.2,10.3) has 1
bin [10.3,10.4) has 1
bin [10.7,10.8) has 1
bin [11.0,11.1) has 2
bin [11.2,11.3) has 1
bin [11.3,11.4) has 2
bin [11.6,11.7) has 1
bin [11.9,12.0) has 1
bin [12.2,12.3) has 1
bin [13.0,13.1) has 1
bin [13.2,13.3) has 1
bin [13.3,13.4) has 1
bin [13.6,13.7) has 1
bin [14.0,14.1) has 1
bin [14.2,14.3) has 2
bin [14.3,14.4) has 2
bin [14.5,14.6) has 1
bin [15.2,15.3) has 1
bin [15.3,15.4) has 1
bin [16.6,16.7) has 1
bin [16.7,16.8) has 1
bin [17.1,17.2) has 1
bin [18.4,18.5) has 1
bin [20.1,20.2) has 1
bin [22.4,22.5) has 1
bin [23.0,23.1) has 1
bin [24.7,24.8) has 1
bin [24.9,25.0) has 1
bin [27.5,27.6) has 1
bin [28.3,28.4) has 1
bin [28.4,28.5) has 1
bin [28.8,28.9) has 1
bin [31.3,31.4) has 1
bin [32.7,32.8) has 1
bin [36.4,36.5) has 1
bin [39.0,39.1) has 1
average edge length histogram 0.4 for 277888 element(s)
--
Download LAStools at
http://lastools.org
http://rapidlasso.com
Be social with LAStools at
http://facebook.com/LAStools
http://twitter.com/LAStools
http://linkedin.com/groups/LAStools-4408378
Manage your settings at
http://groups.google.com/group/lastools/subscribe
Maeva
Jul 9, 2018, 4:16:18 PM7/9/18
to last...@googlegroups.com
Dear Martin,
thank you for your complete answer to this matter :)
I will then generate a 0.8 or 0.5 step raster resolution for my DEMs.
The option "thin with grid" does not have any sense if I understood well, so the final command will be:
C:\LAStools\bin>las2dem -i tiles_classified\*.laz ^
-keep_class 2^
-extra_pass^
-use_tile_bb^
-odir tiles_dtm -obil^
-cores 4
without the thin with grid option.
warm regards!
Reply all
Reply to author
Forward
0 new messages