Resample

images/download/attachments/113546888/CmdCloudRegularSampling.svg

This command allows to keep points regularly spaced on cloud(s).

Select at least one cloud and launch the command.

images/download/attachments/113546888/image2020-3-4_11-25-2.png

Set the value for Average distance between points. This value is used to define the size of the grid cells projected on the point cloud. In each cell, a small part of the surface is computed and the closest point to the surface is considered as the best one and is kept.

Locally, if the density of points is not sufficient to calculate a surface, the grid spacing is increased to an optional size that can be defined by the parameter Max distance between points. If no maximum size is set, the cells can grow to the size of the cloud itself.

Validate the command with OK, display the result with Preview or close the dialog box with Cancel.

This type of filtering is useful to achieve a mesh with a reasonable number of triangles, a good level of details, in a reasonable amount of time when the input point cloud is very large. In this case, you can make a mesh in four steps:

Creation of a rough mesh with 3D Mesh command using the original cloud.
Computation of a point cloud with a space between the points corresponding to the desired level of detail.
Refining with Refine Mesh from Cloud Interpolation, using this reduced cloud, with the option take points from the cloud without noise reduction already removed. The deviation parameter corresponds to the desired level of details.
Second refining but this time using the original point cloud and the option interpolation of new points with a deviation parameter.

Comparison between "Resample" and "Reduce"

	Reduce	Resample
Reduction factor (ratio between the number of input points and output points)	The reduction factor is "under control" and can be low.	The reduction factor is not known in advance.
Noise measurement elimination	No measurement noise removal. Points in low density areas are all preserved unlike the high density areas where points are primarily destroyed.	If a maximum distance is fixed, the points in low density areas are destroyed, because they are suspected of being noisy points.
Point selection	No point is "qualified". The processing is purely statistical.	Each point is singly "qualified" to keep only the best ones.

Note

Limit objects manipulation is available in this command with CTRL+SPACE shortcut.