Smoothing, Point Cloud

Requires an LP360 Geospatial Advanced, LP360 Drone, or LP360 Land Standard license.

The Smoothing, Point Cloud Point Cloud Task (PCT) is used for smoothing noisy point cloud (LAS) data. This PCT is a moving least square smoothing algorithm that reduces “noise” in high density, noisy point cloud data such as that produced by automotive grade LIDAR scanners (e.g. Velodyne, Quanergy).  This smoothing algorithm is notable in that it can be applied to unfiltered data (e.g. non-ground classified data).  While it causes a bit of smoothing of sharp edges, it will not destroy complex vertical objects such as trees. This means that, in many cases, it can simply run across the entire input data set. 

Input Dataset

Select the data to use in the LAS Smoothing process.

• Feature Geometry - Defines the area to apply the smoothing.
• Source Points - Select the points within the dataset to use in the processing.
  • Input Points Filter - Opens Live View to allow filtering of which points to use from the input LAS Layer.

LAS Output

• Output Location - Select an Output location to write the smoothed LAS dataset.

Tip: Using the <> icon found on the right hand side of the I/O Manager will add the LP360 Project Path Variable (<LP360_PROJECT_PATH>) which will automatically output into the currently open Project folder.

• Output Layer Name: Introduced in 2025.2, the user may specify the name for the smoothed LAS Layer, for example, "Smoothed LAS", in place of the default name, "<Default>", where the legacy auto-naming convention is used, "LAS_Layer#".
• Output Structure
  • Tiled Files - Recommended. Automatically generate LAS files of no more than the specified File Size in MB that was introduced in patch A for v2025.2. The default 256MB is the value used prior to the new settings and is the recommended value when using QVR pyramids. When using LPLAS pyramids you may increase the size though shouldn't exceed 4048MB.
  • Single File - Will generate only a single LAS File. Not Recommended unless working with small datasets and LPLAS pyramids.
  • Separately Process Source Files - When this option is used, the source files are processed independently of one another.  This means that, in overlap regions, the data from different flight lines (if it is in flight line structure) will be independently processed.  In general, this exacerbates any vertical shift between flight lines.  On the other hand, if this option is not checked, the data are processed as mixed tiles.  In this case, the separation between flight lines is reduced.  
• Output Points
  • Output All Points - Output all points from the input LAS layer even if they don't satisfy the Input Points Filter and therefor did not get smoothed.
  • Output Only Smoothed Points - Outputs only the points which satisfy the Input Points Filter. 
• Smoothing Point Cloud
  • *This section was moved from the Project Settings -> EXP tab in Patch C for v2025.2.
  • There is an option to smooth 3D or Vertical Only. 3D is useful for photogrammetric, aka dense image matching (DIM), point clouds, as well as for point clouds collected using a handheld laser scanner.
  • The Polynomial Order will increase the smoothing but take more time to compute, the difference is nearly indiscernible on normal datasets both in terms of results and execution time. 

Examples

How to use Smoothing, Point Cloud PCT

Learn more background details and explanation on how the Smoothing, Point Cloud PCT works.