Browse all publications by topic
Browse all publications by year
- A. Kuzmanovski,
M. Stellmacher, and K. Obermayer. A New Algorithm for the Evaluation of
PTV Data Using Point Matching Distance Mesaures and Deterministic
Annealing.
.
In Proc. Int. Symp. on Appl. of Laser Techniques to Fluid
Mechanics, pages 10.4.1-10.4.6, 1998.
(FTP Gzipped PostScript, 6 pages, 203 kb)
We describe a new method for the evaluation of single pulse
multiple exposure PTV data which allows the joint estimate of several local
flow field parameters, among them components describing local translation,
rotation and shear. The method operates on particle coordinates, which can be
extracted from digital images using the CLEAN algorithm (Schwarz 1978), and
is based on an algorithm recently proposed by Gold et al. (Gold, Rangarajan
and Mjolsness, Neur. Comp. 8, 1996) for matching problems in stereo vision
and statistical pattern recognition. For flows with small velocity gradients
our method compares favorable to crosscorrelation techniques. When velocity
gradients become larger, for example near vortices and in shear flows, the
new method leads to strongly improved velocity estimates, and additionally
allows an estimate of the strength of the local rotation and shear
components. The method can in principle be extended to stereo data, because
computational expenses grow linearly with the the number of particle
coordinates only.
|
