Thinning by characteristic relief points (identification of elevations and low points)

[iden]

Hello! My task is to thin out the cloud, but taking into account the characteristic relief points. Conventional thinning does not achieve results when there are more points between a given distance (for example, there must also be the highest in a given square or the lowest). I just started studying the program. Maybe there is already a tool that does this? If a similar topic has already been posted on the forum, please send a link. Thanks for the answers.

[daniel]

Not sure that it will help (also because the answer arrives quite late) but you can subsample the cloud based on a scalar field: https://www.cloudcompare.org/doc/wiki/i ... F_(option)

[iden]

Daniel, yes. According to the tool's description, this is what we need. Thank you, I will try it

[iden]

Please explain what the spacing value sub sampling of the scalar field is responsible for. It's not entirely clear from the link what they're responsible for. After trying some options, it was not possible to achieve the desired result, as per the link guide. Do you have any experience with such sub sampling? Is it possible to perform thinning on the roughness field?

[daniel]

It's a linear model that maps a given sampling value 's1' (e.g. the minimum distance between 2 points if you have selected the spatial mode as in the screenshot) to the minimum scalar field value, and another sampling value 's2' to the maximum scalar field value.

For any point with scalar field value 'v', the sampling value will be s1 + (v - vMin) / (vMax - vMin) * (s2 - s1).

If you have a scalar field that goes from 0.0 to 10.0, and you assign a minimum distance of 50 for sf = 0.0, and a minimum distance of 5 for sf = 10.0, then the minimum distance will be 50 for sf = 0.0, 5 + 45 / 10 = 9.5 for sf = 1.0, 14 for sf = 2.0, ..., 50 for sf = 10.0