Blast2DEM failure

[Crystal Childress]

Blast2dem is failing on "create a tif" for various las files with the following error: FATAL ERROR: vertex not in hash. need pre-order mesh

 

It creates a 1kb tif and does not return an error code.

 

I have attached the Readme.txt that specifies the command line, as well as the 1KB tif, but the laz is 71.6MB, so let me know the best way to get that to you.

Thanks,

Crystal

[Jochen Rapidlasso]

Hi Crystal,

please do a lasinfo64 -i ... of your file and send the console output.

Does the command work if you use LAZ as output (-olaz)?

If you want to provide your input file for analysis its best you put it on a cloud storage and send the link.
If its public data you post the link here, otherwise please send a personal/private message/mail.

If you do not have any options to get a cloud space please write us a note and we will give you a space on our server...

Best,

Jochen @rapidlasso

[Crystal Childress]

-olaz gives me the same error.  The lasinfo.asc is attached.  

[Jochen Rapidlasso]

Hi Crystal,
we asked for the output of lasinfo64 to get a glue how your data looks like.
To produce an output file with lasinfo64 you can use the command
     lasinfo64 -i your_input_file.laz > lasinfo.txt 2>&1

Then send the lasinfo.txt file to us.

Anyway, best if you send the full laz file by any fileshare/cloudspace.

Best,

Jochen @rapidlasso

[Crystal Childress]

Apologies, Jochen.  The lasinfo is attached.  

[Dipro Sarkar]

Hi everyone,

I too faced this problem yesterday. When I retraced my steps, I found that using lasthin with '-percentile' flag while thinning the cloud causes this error for me in blast2dem. When I started using any other method for thinning and then using the thinned cloud in blast2dem, the error vanishes. 

I hope this help. 

Joe

--
Download LAStools at
https://rapidlasso.de
Manage your settings at
https://groups.google.com/g/lastools/membership
---
You received this message because you are subscribed to the Google Groups "LAStools - efficient tools for LiDAR processing" group.
To unsubscribe from this group and stop receiving emails from it, send an email to lastools+u...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/lastools/89ced158-cd9c-4872-8e18-756915a32627n%40googlegroups.com.

[Jochen Rapidlasso]

Hi Crystal,

thanks for your data and sorry for late reply.
What you want to do is a kind of a challenge:
You have a input of 804582 x 743943 points and want to create a TIF of the size 805 x 745 (-step 1).
The input file looks quite raw - with a some noise and artifacts below and above the area where we expect data.
Also you did not do a ground classification and no height calculation.
Additional you have a large offset in your header.
 
We tried to remove each challenge one by one to see if this is the cause for the error calculating the output.
Finally it seems to be a numerical problem caused by the combination of those challenges, following the 1st rule of IT systems:
If you put crap in, you get crap out.
blast2dem  triangulate your data and calculate the normals at the non-ground-points location to define a output value/color for the according output grid point.
If there are combined (large_numbers + unrealistic_slopes) it is not possible to calculate useful normals.

Fortunately there are several ways to avoid this situations.
One workaround can be just to translate the data to lower values, meaning reduce the offset value of 4*10^6 - we did tests with
        copy test.laz tmp_offs.laz
        lasinfo64 -i tmp_offs.laz -set_offset 0 0 0 -repair_bb
        lasinfo64 -i tmp_offs.laz
        blast2dem -i tmp_offs.laz -step 1 -meter -slope -last_only -nodata 255 -o tmp2.laz
which works well.

This was just for testing, much better is just to remove the noise - it sure doesn't make sense to include those far_outside_points into a strange DEM:
    lasnoise64 -i test.laz -remove_noise -o tmp1.laz
    blast2dem.exe -i tmp1.laz -step 1 -meter -slope -last_only -nodata 255 -o tmp2.laz
This removes (just) 613 noise points out of your 21000001 points and blast2dem will be able to produce a useful output.

Best regards,

Jochen @rapidlasso

[Crystal Childress]

Thank you so much for the fast response time. It’s not surprising this data might be causing issues. It seems to work 99+% of the time, so we were hoping there was something obviously wrong with this data. The successful files are very similar (just running lassort first seems to fix some of these). 

 

We will see if doing some noise filtering up front is viable for our workflow.  The rasters we produce are used for initial QC to determine if the las created meets spec or needs a re-flight, so we try not to do too much manipulation.

 

As for blast2dem, could it have more meaningful error messaging? I know you are unsure what exactly is causing the problem, so that might not be possible.  At the very least it would be nice if it returned an error code so the failure could be caught earlier.

To view this discussion on the web visit https://groups.google.com/d/msgid/lastools/6cda486d-21f5-4bef-a6be-8eb4fb03b4efn%40googlegroups.com.