For Full-Text PDF, please login, if you are a member of IEICE,|
or go to Pay Per View on menu list, if you are a nonmember of IEICE.
A Neural Network Based Algorithm for Particle Pairing Problem of PIV Measurements
Achyut SAPKOTA Kazuo OHMI
IEICE TRANSACTIONS on Information and Systems
Publication Date: 2009/02/01
Online ISSN: 1745-1361
Print ISSN: 0916-8532
Type of Manuscript: PAPER
Category: Image Recognition, Computer Vision
flow measurement, particle image velocimetry, particle tracking velocimery, neural network, optimization,
Full Text: PDF>>
Particle Image Velocimetry (PIV) is a widely used tool for the measurement of the different kinematic properties of the fluid flow. In this measurement technique, a pulsed laser light sheet is used to illuminate a flow field seeded with tracer particles and at each instance of illumination, the positions of the particles are recorded on digital CCD cameras. The resulting two camera frames can then be processed by various techniques to obtain the velocity vectors. One such techniques involve the tracking of the individual particles so as to identify the displacement of the every particles present in the flow field. The displacement of individual particles thus determined gives the velocity information if divided by known time interval. The accuracy as well as efficiency of such measurement systems depend upon the reliability of the algorithms to track those particles. In the present work, a cellular neural network based algorithm has been proposed. Performance test has been carried out using the standard flow images. It performs well in comparison to the existing algorithms in terms of reliability, accuracy and processing time.