Hi Thiango,
That is a large difference.
I see you aren't building -nurbs- faces from the STL data, so that difference
in memory consumption is strange.
Have you tried various .stl files?
The difference in performance can perhaps be partially explained on a
functional basis. PythonOCC deals with CAD, essentially higher order nurbs
geometry, while meshlab is highly optimized for massive polygon data. So, in
that sense it is not unexpected to see a performance difference ( albeit not so
large ).
PythonOCC can be most useful for say, fitting nurbs through your STL pointcloud.
However, for dealing exclusively with polygonal data, Meshlab is the way to go.
Best,
-jelle
On Jun 24, 2010, at 3:13 PM, Thiago Franco Moraes wrote:
> Hi all,
>
> I've been trying to read STL files. With small STL files it reads
> normally. With not so big STL files (like the one I've just tried to
> read that have 51 MB, binary, 534678 vertices, 1069782 faces) it can't
> read, the CPU usage is about 100% and memory grows up from a few MB to
> 4 GB then I have to kill my script. Using Meshlab the memory usage is
> about 270 MB. I've tried to use STLImporter class and StlAPI class.
> Bellow the code I used:
>
> from OCC.Utils.DataExchange import STL
> reader = STL.STLImporter('file.stl')
> reader.ReadFile()
>
> Ah, I'm using the packages from Ubuntu 10.04 64 bits, the last version
> in PPA repository [1].
>
> I have done something wrong on trying to read that file? Is it a bug?
> If necessary I can upload the STL in some place.
>
> Thanks!
>
> [1] - https://launchpad.net/~cae-team/+archive/ppa
>
> _______________________________________________
> Pythonocc-users mailing list
> Pythonocc-users@gna.org
> https://mail.gna.org/listinfo/pythonocc-users
_______________________________________________
Pythonocc-users mailing list
Pythonocc-users@gna.org
https://mail.gna.org/listinfo/pythonocc-users
