🚀 go-pugleaf

In response to my posting (article 24902@Princeton.EDU):

>>I'm working with VHS video of a complex water flow seeded with particles 
>>and I'm trying to track the particles' paths.  >

In article 15769@chemeng.ed.ac.uk, andreww@Chemeng.ED.Ac.Uk (Andrew K.
Hind) writes:

>I know of several people working in this field using a variety of
>techniques. It would help to know if you want the whole filed or just a
>single plane. I assume you are going for the plane.

I am currently involved in planar flows, however, in the future I will
be looking at 3-D flows, so any information on PIV techniques in 3-D flows
would be useful as well.

>I know of several ways to obtain the veocity vectors, it depsends on
>the seeding density which technique will be appropriate (fast and
>accurate).

>Particle tracking - essentially low seeding density.  Particle image
>velocimetry - medium seeding density Speckle velocimetry - high seeding
>density

My flows by necessity have low seeding densities.  I have tried
techniques used on high density flows (cross correlations, etc.) but
with no luck :-(.

The papers that I have found so far (I haven't yet looked at the
references in the posting) seem to use particle traking algorithms that have 2
qualities: 1) they involve lots of computations which are usually done
on supercomputers, 2) they involve lots of "user adjustable" parameters
that are arbitrarily chosen.  I am hoping to avoid both these qualities
in my analysis (as much as possible).

Much thanks to Andrew Hind for responding so extensively.

Arel Weisberg
weisberg@phoenix.princeton.edu