Workflow to process LiDAR

[Craig Unpronounceable]

Martin,

Wanting to process our LiDAR data in the most efficient manner
possible to create a hillshaded image of the DEM, contours (not yet
included) and vegetation. For the vegetation I am interested in trying
to extract the areas which are "dense" which has a canopy height
between 1 and 3m (which makes it difficult to run through).
The LiDAR data we have access to is currently in 1kmx1km tiles. I have
created the (very simple) batch file below to process all the files in
a directory and I would appreciate your feedback on whether the logic
is correct ie lasground before lasmerge etc.

***********
REM extract ground points
lasground -i *.las -olaz -v -cores 7

REM merge all files together
lasmerge -i *.laz -o merged.laz -v

REM create a DEM and output an image
las2dem -i merged.laz -keep_class 2 -opng dem.png -hillshade -v

REM height DEMs and output images
lasheight -i merged.laz -o height.laz -replace_z -v
lasclassify -i height.laz -o classified.laz -v
las2dem -i classified.laz -o classified.png -false -v
***********

Thanks in advance.

[Terje Mathisen]

Craig Unpronounceable wrote:
> Martin,
>
> Wanting to process our LiDAR data in the most efficient manner
> possible to create a hillshaded image of the DEM, contours (not yet
> included) and vegetation. For the vegetation I am interested in trying
> to extract the areas which are "dense" which has a canopy height
> between 1 and 3m (which makes it difficult to run through).

Another orienteering mapper!
:-)

There is one little tweak that I've been working on, and that is to try
to locate areas with high canopy (i.e. tall trees, usually quite
runnable), but where there is a significant amount of reflections in the
same 1-3 m range, since this means dense under-vegetation.

Effectively what I want to do is to filter away all vegetation above
maybe 5 m, except that the expected density of 1-3 m reflections in an
area with 10+ m forest is significantly lower that for an area with
nothing but brush.

> The LiDAR data we have access to is currently in 1kmx1km tiles. I have
> created the (very simple) batch file below to process all the files in
> a directory and I would appreciate your feedback on whether the logic
> is correct ie lasground before lasmerge etc.

There are some limits on how large an area you can run contour
generation on, but the best approach is probably to use Martin's
streaming tools, since they do the work in 256x256 cell tiles, with no
risk of running out of memory in the middle of the operation.

>
> ***********
> REM extract ground points
> lasground -i *.las -olaz -v -cores 7
>
> REM merge all files together
> lasmerge -i *.laz -o merged.laz -v
>
> REM create a DEM and output an image
> las2dem -i merged.laz -keep_class 2 -opng dem.png -hillshade -v
>
> REM height DEMs and output images
> lasheight -i merged.laz -o height.laz -replace_z -v
> lasclassify -i height.laz -o classified.laz -v
> las2dem -i classified.laz -o classified.png -false -v

When I did something similar for the area where we intend to host the
Junior WOC in 2015, I ran lasground/lasheight/lasclassify against
individual files (my source is in 1200x1600m tiles, i.e. a little less
than 2 sq km for each file).

I used a grey scale output from lasclassify, with fixed limits of 0.0 to
25.5m, so that each step in the 0-255 byte range corresponds to 0.1 m.

I then ran this image through photoshop (could have used gimp just as
easily):

First I converted the grey scale image into an indexed color format,
this allowed me to manually edit the palette so that various heights
would be mapped into various color ranges.

Finally I converted this back to RGB and saved it as high-quality JPEG.

Even without the "intermediate vegetation" tweak this is very useful,
and the current colors match up very well with mountain birch vs spruce
and pine areas.

Terje
--
- <Terje.M...@tmsw.no>
"almost all programming can be viewed as an exercise in caching"

[Craig Unpronounceable]

LiDAR is great for ground features but like you Terje I also want to
extract vegetation features. Hopefully Martin will have some
suggestions :)

Thanks for the other pointers about getting base layers prepared.

Craig

> - <Terje.Mathi...@tmsw.no>