CloudCompare forum

Hi Daniel,

With the latest version of our 3D point cloud comparison software we output ascii files (boouuh) and the first line contains the name of the columns (and there are several column). When I skip this line, Cloudcompare guess correctly that the first 3 columns are coordinates, and the 3 next normals, and the others scalars, but it'd be great to have the name of the scalar fields read directly from the header line (rather than having scalar #1 etc...).
The first line of our files read as :
#c1.x c1.y c1.z n1.x n1.y n1.z diff conf_interval etc...

Alternatively, is there a particular header format that Cloudcompare can use to load the name of scalar fields in an ascii file ?

Other question : is there a keyboard shortcut to scroll through the various scalars ? (like arrow up and arrow down ?)

Cheers and congrats for the latest version of CC, it's great !

Dim

I like the idea of adding shortcuts for changing the current scalar field! (and it's a very easy request... for once ;).

For the ASCII format, there's no predefined format (the idea is to be the most generic as possible). But we could agree that if only the first line is skipped and that this line has exactly as many elements as the number of columns, then we could use this information for naming scalar fields.

This means that either you don't use any space character in the scalar field names or that you use a different separator for the whole file (semicolon, etc.).

HI Daniel,

that'd be perfect...I'm used to never use column names without space character.

Cheers


Dim

I hope you meant the opposite ;)

It's already done on the trunk. I'll try to issue a new release at the end of the week.

oups, I'm a bit tired... currently rushing to finish the paper on the 3D point cloud comparison algorithm before the holidays....

Hi guys,

What is the format of the header for an ASCII file type? eg "x y z s r g b"?

Does having an ASCII header line mean that the "open ascii file" dialog will be skipped?

Many thanks, Matt

Nope ;)

For the moment only the scalar field names can be extracted from the header line. To be valid this header line must contains as many "words" as columns in the file (so you can't use space characters or equivalent delimiter in your column names).

By the way, Daniel, thanks for this functionnality. This is ultra useful !