Dynamic Open Contours Using Particle Swarm Optimization with Application to Fluid Interface Extraction

Abstract. This paper describes a method for the estimation of a dynamic open contour by incorporating a modified particle swarm optimization technique. This scheme has been applied to a "Particle Image Velocimetry" experiment for the analysis of fluid turbulence during a hydraulic jump. Due to inter reflections within the medium and refractions across different media interfaces, the imagery contains spurious regions, which have to be eliminated prior to the estimation of turbulence statistics at the fluid surface. The PIV image sequences provide a strict test bed for the performance analysis of this estimation mechanism due to the occurrence of intense specularity and extreme non-rigid motion dynamics.

• Type: Conference
• Year: 2006
• Where: ACCV
• Authors: Mani Thomas, S. K. Misra, Chandra Kambhamettu, J. T. Kirby
• Keywords: ACCV 2006, Computer Vision, Estimation, Particle Image Velocimetry, Particle Swarm Optimization