Min num of total sample points for tricubic convolution?
- From: robert <robert.kj.karlsson@xxxxxxxxx>
- Date: Thu, 20 Mar 2008 15:14:30 -0700 (PDT)
I have a volume filled with scattered sample data (sort of like a body
of water with measured temperatures for different locations). I am
planning on using interpolation to get a *rough* estimate of the
values that fall in-between the sample data, and was thinking about
(tri)cubic convolution.
However, since that would require a grid of 4x4x4 sample points (64)
for the calculation of an in-between point, it would come dangerously
close to the actual total amount of sample points, which are 200-300.
Would this mean I am better of using something simpler, like trilinear
interpolation?
(forgive me if this is a naive question - I'm pushing my rusty upper
secondary school math to the limit!)
Robert Karlsson
.
- Prev by Date: math -- boolean combinations of rectangles
- Next by Date: Re: Cylinder to Cylinder Intersection - Two-Dimensional Shape
- Previous by thread: math -- boolean combinations of rectangles
- Next by thread: math : a challenge from tommy1729
- Index(es):
Relevant Pages
|
