Hi,

Is there a way to create a surface from any arbitrary number of points? I've got about a million points being imported and want to create polys and a NURBs surface from them (in the HDK) but was wondering if there's a SOP to do that already. It's probably more efficient
to do it with HDK but was just wondering …


–Mark

An Add SOP can turn a set of points into Polygons and from there you can convert them into a NURBs surface.

However getting the point ordering and such wouldn't be the most fun thing in the world.

Another alternative would be to convert the dataset into a geo file and read that in. (Or a unix sop or something…)

But like you said..its probably more efficient to do it in the HDK.

I think CySlice has a good facility for point clouds. You could take a look at that maybe?

What type of surface are we talking about? For complicated geometry, there's many relevant hits via a google search of “point cloud algorithm”. For terrain, the povray discussion led to:
http://astronomy.swin.edu.au/~pbourke/terrain/triangulate/ [astronomy.swin.edu.au]

For fun though, you could try converting the point cloud into an .i3d file via i3dconvert and bringing it back into an isosurface sop.

Wolfwood
An Add SOP can turn a set of points into Polygons and from there you can convert them into a NURBs surface.

However getting the point ordering and such wouldn't be the most fun thing in the world.

Another alternative would be to convert the dataset into a geo file and read that in. (Or a unix sop or something…)

But like you said..its probably more efficient to do it in the HDK.

Hi there,

Yes, I thought the Add SOP might be able to do something like this, I'm trying it now but it doesn't appear to be the solution I'm looking for (I'm getting undesirable artifacts). Another idea I had was like you said to write a simple program to spew out a RIB file and then read that in … but now that I think about it, it may be easier to generate a .geo file.

This data are the topograhic data of the surface of Mars, and I'm using the lowest resolution data set available, 4 samples per degree (720 X 1440) which is 1,036,800 points. Eventually I'll want to use 16/32/64/128 samples per degree …


–Mark

jason_iversen
I think CySlice has a good facility for point clouds. You could take a look at that maybe?

Hmmm … not sure, I took a look at the web site (having never used it before) but I don't know if it'll take a raw set of points, I'll see if I can find more.

Thanks for the suggestion.


–Mark

edward
What type of surface are we talking about? For complicated geometry, there's many relevant hits via a google search of “point cloud algorithm”. For terrain, the povray discussion led to:
http://astronomy.swin.edu.au/~pbourke/terrain/triangulate/ [astronomy.swin.edu.au]

For fun though, you could try converting the point cloud into an .i3d file via i3dconvert and bringing it back into an isosurface sop.

Wow, now this looks interesting! I was just going to stitch together “poly quads” but had a gut feeling that wasn't a good way to approach the problem.

For grins I just may try to write out an i3D file.

Thanks!


–Mark