[Gmsh] Help

Henri Pascal-Jenny henri.pascal-jenny at sogeti.com
Thu Jun 17 11:26:36 CEST 2010



I am facing the  problem of finding the minimum-3D box enclosing AS CLOSE AS
POSSIBLE a set of 3D mesh points . 


I would like to avoid the use of a too-sophisticated algorithm .

(cf.   http://en.wikipedia.org/wiki/Minimum_bounding_box_algorithms)


Actually, when used straightforwardly, the method "BoundingBox"  available
in a  gmesh-session returns a  3D box  

which box does not always correspond to the minimum volume enclosing the set
of points 


( imagine a thin and not fully-planar shape tilted in respect to the axes of
a standard coordinates-frame )


In fact, the matter comes from the bulk orientation of the envelop of the
set of points under consideration.


The usage is : 

BoundingBox { expression, expression, expression, expression, expression,
expression };

what forces the bounding box of the scene to the given expressions (X min, X

Y min, Y max, Z min, Z max).


Therefore, I am wandering if there is a way of  operating the "BoundingBox"
method in order to get in the end the  truly-minimum 3D box enclosing. 

Actually, I guess that there is a preliminary thing to do related to the
coordinates system ; maybe I am wrong.


Well, I would appreciate it very much if anyone could supply me for any
piece of information regarding my problem. 

Thanks a lot in advance for your attention and your cooperation.




Henri Pascal-Jenny 



-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.geuz.org/pipermail/gmsh/attachments/20100617/9522a6bb/attachment.html>