regular 2.5D surface raster

[Alex B]

Dear Daniel,

I have an irregularly dense TLS point cloud (high-res, with a mean point spacing of 3mm) and would like to get a regular 2.5D-spaced raster of the scanned surface - like if an evenly-spaced flat grid (here 1mm) is put on top of the meshed points and the xyz-values of the now-folded grid are taken.

So far, I tried the rasterize tool, and I get a homogeneously spaced grid as seen from a projection direction (e.g., Z). However, I wonder, if it is somehow possible to get an evenly spaced grid on the actual topography? Maybe I just missed the simple way to do that or didn't find the right tool ...

Cheers and thanks,
Alex

[daniel]

Do you mean that you wish the geodesic distance between points would be regular? As if you were 'texturing' the ground with a regular pattern? In this case I fear there is no solution in CloudCompare indeed...

[Alex B]

Hi Daniel,

thank you for the help! Yes, that is what I meant. Okay, in this case I must stay with the rasterized version.

Then, another question :): As I have a curved, open surface, by the rasterize tool I get points that actually lay "outside" due to the underlying assumed interpolation rectangle of the tool that partly exceeds it. Since I cannot cut all these undesired points by hand, I tried to mesh the original cloud first and then use the "sample points" tool with a given density. Would this tool give me kind of the desired grid?

That works with a defined number of points (with irregular spacing), however, with given density, I always receive an error : "[doActionSamplePoints] Errors occured during the process! Results may be incomplete!".

Best,
Alex

[daniel]

Do you need the 'interpolation' in the Rasterize tool? Because it's optional (you can leave the empty cells empty.

And for your error with the sampling tool, don't you have another warning message before the error message in the console? Because generally if you input a too small density, no point will be generated. Otherwise I would be interested to know more about this issue!

[Alex B]

Thanks again Daniel!

A. rasterize tool: Yes, I need it, since my given point spacing is lower than the aspired one - so I want to interpolate. This then comes with interpolation at "outside regions", where I don't want it :).

B. sample point tool: Yes, I get another error: [ccGenericMesh::samplePoints] No point was generated (sampling density is too low?)
Ah, I misunderstood the tool's definition - I thought it was meant to enter the resolution, not the amount of points for the density ... So now, with both options (points number or density) I get 1M points per m2, corresponding to a mean point distance of 1mm. However, looking in close-up, their spacing is much more irregular than with the rasterized version ... How is their position actually calculated? Would it be possible to have a kind of "rasterized version" here, too? This then would correspond to my original question (if I cut the mesh accordingly before).

Cheers,
Alex

[daniel]

A. OK

B. The 'Mesh sampling' tool is just randomly sampling tool on the triangles. Hence the irregular spacing... And no there's no 'regular' sampling method for meshes.

[Alex B]

Okay, thanks you for replying - I'll try the rasterized version then