Problem with output file

Added by Teuber Joseph about 2 years ago

Hello Dr. Hackenberg,

first of all I would like to thank you for the ongoing effort and support that you provide so that others can use your software.

I am using CompuTree to investigate the herbivory effect on South African savanna trees with TLS-data. At the moment, I am trying to get a better understanding of the LiDAR-technique and the CompuTree-Software by practicing in local German forests, before I will go to South Africa in February.
In doing so, some problems came up regarding the export of the data from the QSM-models from CompuTree for which I could not find any solution:
I followed all of your instructions from your youtube-tutorials and I think my QSM´s are looking good. But when I am exporting these data, the volume of each single tree (and also of the cylinders in the detailed file) is way too little (values like 0,0234 m³). I loaded the .ply-data from the output file of the single trees in CloudCompare and saw that many cylinders that are shown within CompuTree are missing when visualised by CloudCompare. So I think the problem accures during the export of the QSM-data from CompuTree and many cylindres are lost during the export. Attached please find a little illustration of my problem showing my QSMs from CompuTree (4 trees) and the exported QSM-output from them in CloudCompare.

Have you ever had to deal with this problem or do you have any solution or suggestion to solve my problem?

Again, I am very grateful for any advice!

Best regards from University of Bayreuth,

Joseph Teuber

Replies (1)

RE: Problem with output file - Added by Hackenberg Jan about 2 years ago

Hi Jospeh

first of all thanks for your interest in SimpleTree. In fact you asked two questions:

1) The missing volume. I never heard of it, and obviously I did not design such behaviour by intention. There is two pieces of information missing in your description. You write missing volume. That means you have a volume to compare to. Do you mean a destructive ground truth volume? Every measurement has an error - thats rule number one in statistics. So your ground truth has an error, and obviously the QSMs have an error as well. In that case the 0.0234 is a combined error. That number looks really small though. So the second option is that you compare my total volume with something like the summed up volume of the detailed files. There might be rounding issues. Floating points (non natural numbers) computation are inaccurate by the design of computers. Computers can just approximate. If one volume then is computed with C++, the other with R or Excel, here might be an explanation. But I would be still shocked, cause in that case the error is quite large. The third option is that this is caused by a bug of my plugin. Most times bugs do out of the line errors or crashes, but I recently found another bug which made it through.... Maybe this already helps, otherwise please give me more information here.

2) The second issue of the missing cylinders is relatively simple. The QSMs contain two kinds of cylinders. The first group is cylinders purely fitted with geometric routines. The second group is the cylinders fitted in same manner, but have a statistical corrected radius. The reason is, that I expect users are interested in doing their own statistics sometimes. Those researchers need to know which cylinders they can use to build their own statistic (first group) and which not. If they find better routines, they can adjust the radii of group 2 in their own manner. See here:
Full QSMs, then only less cylinders in the following seconds. That is intended. I remember though that I produce multiple ply files. Please check your output folder. It should contain various sub folders for plys. Check the other folders or give me a warning if there is no such folders.

I hope that helps