How to reduce a Lidar point cloud?

NicolasPirotte

Hi, I have a Lidar point cloud of an UG mine and the file contains too many points to allow creation of a solid in Leapfrog (an intrusion). I remember having seen a way to select every 10th, or 50th..., point from a list. Could you remind me how to create such filter?

JamesEdwards

Hi Nicolas, 

You can use a query filter to remove a % of points. It is worth noting this is a random selection.

[id] % 5 = 0

In this case, the query is saying use every 5th point (keeps 20% of data)

[id] % 5 != 0

In this case, the query is saying remove every 5th point (keeps 80% of data)

If it is a topographic survey you can use the error threshold in the triangulated mesh to set a tolerance instead which will give a better result.

MichaelDeVasto1

Nicolas, let us know if the above worked for your needs. 
As I'm reading this I'm thinking that even if you're able to reduce the number of points in your point cloud, you will not be able to create the instrusion as intended. 

For example - if you take an existing solid in LF (leapfrog) and extract verticies (i.e. your point cloud), and attempt to remake that intrusion from those points, you'll not be able to duplicate this. Singular points with XYZ and RBG doesn't tell leapfrog if they are contact points - the method used to build intrusions in LF. 
I'm unsure how to complete this task in leapfrog - I would suggest working with whomever supplied the point cloud to generate a mesh that represents the tunnel and import that into LF and create an intrusion from there. 

I've had pointclouds from surface LiDAR that I've attempted to surface and it has poor results, simply too many points that made an unusable mesh (misc. points "off surface" created many high and low points). To work around this, I'd with a GIS to create the surface for you however, a pointcloud generated from a tunnel may not work so well and would require some specialized help.