Within n4ce 4.40 and above we have now introduced two new features within the point cloud engine called:
Move Noise to Group
Use the Clipping box --> Test parameters on a small number of points --> Remove Clipping Box and apply.
The new tool is designed to remove noise or "fuzz" commonly found within Point Cloud scans as this will help to visualize, process, and digitize data such as topographical surveys. When using the tool we recommend first trying the settings on a small area of the overall scan by using the Clipping Box.
Once the data has been clipped to a specific area of the cloud we must now adjust the tools options to best suit the data. The below table explains each property of the tool. Also, the image contains the default settings. The Move Noise to Group tool works by first breaking down the visible points into Clusters which are then checked for a total number of points. If the count is below the Minimum Cluster size then the points are moved and identified as noise.
Note: A new group called e.g. Noise will need to be created first and set as the Target group.
Group by Normals
The normals tool is designed to group points based on the "normal" (angle) of the surface to identify floors, walls, or custom gradients within a scan. This can be extremely useful when trying to extract floorplan or elevation information from point clouds.
Start by importing the cloud and selecting the "Group by Normals" tool. Now you should see the properties panel appear. By default, the tool has Walls and Ground automatically selected but you can always add a custom normal if required.
Once the settings have been adjusted to suit. For example, the larger the Tolerances the more points will I identify as Ground or Walls. We are using this tool to help remove any unwanted data such as chairs, tables, cars, trees, and other objects that are unnecessary for the process we are trying to achieve.