LandXML Processing Tips

This request came in today and I feel that it has a number of interesting aspects to it that are likely not 100% apparent from the outside.

The user was provided with a LandXML file that they said took a “while” to import and that once imported made TBC a little sluggish in use, so I took a look at what was provided.

The file was a LandXML produced by Topcon Magnet Office. The surface model was defined in the correct manner in the LandXML file as a list of Point locations and then a list of 3D faces, however the file also included the 120,000 points of the surface as COGO Points. That may have been intentional maybe not, so when TBC imports the data it has to read and create 120000 points and then read 120000 nodes and faces and create the surface - this of course does take a little time. Secondly once the data has been imported, selecting 120000 points in TBC is pretty slow - point data is notoriously clumsy in TBC when you have a lot of points - you are way better using the point data as a Point cloud if you need the points, as that is way faster to work with even on a great computer.

In the video we show the contents of the original XML file and explain the different sections, we edit the XML file to remove the COGO Points, and then we show the speed of importing the XML with and without the Cogo Points. We also look at the Surface to Points command that allows you to create a Point Cloud from an imported TIN model, that gives you efficient point data that you can work with. We also show how to select and delete the COGO points if they are imported.

While most people will not open a LandXML file and look at its contents, I do that a fair bit to know what I am working with - in this case more data than needed was written into the file and with some quick process steps you can make it way more efficient to use.

Alan

1 Like