Abstract
An automatic particle-tracking algorithm previously developed by the authors is improved for practical use and applied to measurements of velocity fields of thermal convection. The main part of the algorithm consists of the Kalman's filtering theory and the Chai-square test. The presented algorithm employs a generation method of Delaunay network to estimate initial values of velocity vectors of newly-appearing particle images in the source light sheet.
The results of measurements are compared with those by Adrian et al. by means of LDV and a theoretical prediction by the authors. It is shown that particle tracking velocimetry by the presented algorithm is applicable to measurements of flows with heat transfer.
