抄録
Particle tracking velocimetry (PTV) is one of the methods that measure flow velocity fields. This paper proposes a new method that solves the correspondence problem of particle tracking velocimetry using a network model. The PTV offers many advantages for the study of fluid flows. The flow visualization images are analyzed to obtain the velocity distribution of the flow. Several PTV algorithm has been proposed in the past. The most of the method assumes that the fluid motion, within small regions of the flow fields, is parallel over short time intervals. However, actual flow fields may have some detoured motion, such as rotation and expansion. The proposed pairing particles method using the networks model is based on a connection of particles according as the weights between particles on a 1st image and 2nnd image. The technique is applied to the PTV standard images distributed through the web site (http : //www.vsior.jp/piv). The simulation results demonstrate the effectiveness of the present technique.
