very steep high densitiy lidar processing

monett...@gmail.com

unread,

Apr 10, 2019, 8:27:48 AM4/10/19

to LAStools - efficient tools for LiDAR processing

Hello lidar fans,

I`m working on a very steep high density lidar data, collected with a riegl vux1 sys uav laserscanner.

Handling the very rough and sometimes overhanging situation makes it really hard for me to generate a DGM without smoothing the ground to much.

Especially the vegetation in the stairs is challenging me.

What are your thoughts, experiences, recommendations with such situations?

Thank you very much,

Best regards

David

Photos:

1) RGB GroundPoints whole mine

2) Intensitiy Tile Ground and NoGround-Points

3) High RGB Tile - stone blocks are false classified

4) Intensitiy NoGroundPoints -> Way too much false classified points for me!

Jorg

unread,

Apr 10, 2019, 9:08:01 AM4/10/19

to LAStools - efficient tools for LiDAR processing

Hi David,

Interesting problem. I have been faced with a similar problem a while ago. We flew a VUX along a 300km long coastal cliff (mounted on a manned motorglider). The cliffs are near vertical and also have overhangs and even caves at the bottom. They are on average 30m high with a wild ocean at the bottom and nearly totally flat landscape at the top. I tried many avenues to get a good (vertical) ground surface, but I still have not yet found a really satisfying solution.

One approach was to generate a potree from a few kilometres of cliffs using laspublish.exe. This is one of the better visualisation I was able to generate. It is accessible at:

http://www.airborneresearch.com.au/potree/cliffs2-6/cliffs2-6.html

I found that Microsoft Edge is the best browser to look at it, but Firefox or Chrome also work (sort-of). I am not sure if it works properly in other browsers.
Once the image displays, use the top-left corner menu to set the point count to max and the point size to smallest. Then under "Material" set, the Attribute to "Color" or "Intensity". That gives the best rendering (once you zoom in or out or pan around).

I am looking forward to people giving you advice with your problem, because that may then also be useful for me to try with my data. If anybody is interested in the full set of laz-files (and even a 1-hour video showing the cliffs while were scanning them) - that's all available. Just ask. The video is at https://www.dropbox.com/s/d31hod1ec343h9k/NullarborCliffs2016_VIMEO_HD1080P.mp4?dl=0 , but it is 7GB ! It even comes with music, because it is a bit boring to watch, except if one is interested in geomorphology. The geomorphologists I am working with are rather excited about it.

Cheers from Down Under,

Jorg

Mark Levitski

unread,

Apr 10, 2019, 9:28:32 AM4/10/19

to last...@googlegroups.com

For the false ground points, you just need to adjust your settings for the routine. If you are using Lastools, then there is plenty of documentation to show you the way.

As far as overhangs and lidar shadows, flight planning plays a big role because the swath is never just straight down nadir and therefore can pick up some of the near vertical surfaces.

In the end, one cannot count on automatic routines only. Some manual work usually needs to be done.

Good luck!

Mark

Sent from Mail for Windows 10