lasground - Beginner learning to process lidar data

[Cecilia Chavana-Bryant]

Hello everyone,

I am trying to use lasground to process lidar data (from a GeoSLAM sensor) for the first time and I have a couple of questions that I would be very grateful for some help with:

The size of the files I am processing vary from 1 to 4.5KB. The first file I processed using lasground is ~3KB with ~108 million points. My questions are:

1. This lidar data comes from forests, is using "wilderness" the most appropriate setting? Would it make much difference if I used "nature" instead?

2. Should this process be run as "RUN32" or RUN64"? Does this make much difference? I am using a newish Dell laptop

3. The processing of this file took 278.739 secs and I received the following message:

 WARNING: output slightly distorted. tiny xyz noise. points permuted. intensity, gps_time, user_data & point_source_ID zeroed. 

What does this all mean? AND do I need to worry about this?

4. Due to the above warnings, I decided to run lasinfo on my file (I did not collect this data and just by viewing it I can see empty gaps within the scanned plots, as can be seen below, I received quite a few warnings, what kind of problems will arise from these?

Apologies in advance for any beginner shortcomings in this message, this is the first time I process lidar data and use LASTools.

Thank you in advance for your help,

Cecilia

[Martin Isenburg]

Hello,

The size of the files I am processing vary from 1 to 4.5KB. The first file I processed using lasground is ~3KB with ~108 million points. My questions are:

You must mean GB ... not KB.

108 million points is too much for RUN32. You need to RUN64.

 

1. This lidar data comes from forests, is using "wilderness" the most appropriate setting? Would it make much difference if I used "nature" instead?

Either way should work, see which one you like better.

One thing you could do is *not* include every point for ground processing. These two or three pass approaches first select a subset of points of even density and then only run lasground on that subset of points. I could imagine your GeoSLAM data is very dense in some areas and less dense in others so using lasthin to first select a subset of points to run lasground on is recommended. Maybe try an approach similar to the one outlined here:

https://rapidlasso.com/2021/04/08/generating-dtm-for-fluffy-livox-mid-40-lidar-via-median-ground-points/

Working in tiles (as done in this blog article) rather than on the complete 100 million point data set is another recommendation.

3. The processing of this file took 278.739 secs and I received the following message:

 WARNING: output slightly distorted. tiny xyz noise. points permuted. intensity, gps_time, user_data & point_source_ID zeroed. 

What does this all mean? AND do I need to worry about this?

This is part of the license protection of LAStools. For creating a DTM these tiny distortions will hardly matter.

4. Due to the above warnings, I decided to run lasinfo on my file (I did not collect this data and just by viewing it I can see empty gaps within the scanned plots, as can be seen below, I received quite a few warnings, what kind of problems will arise from these?

The WARNINGs you see here will not matter much. It seems the software who wrote the LAS/LAZ file did not take proper care with setting the bounding box, setting the return count and number, and setting the correct bit in global encoding to specify Adjusted Standard GPS time.