Iterative particle matching for three-dimensional particle-tracking velocimetry

Abstract

A new evaluation scheme for double exposure three-dimensional particle-tracking velocimetry is proposed. Its main feature, a robust multi-pass matching algorithm, is presented and validated by investigating its performance when applied to a synthetic data set. To evaluate real measurement data, the approach is supplemented by an iterative triangulation scheme, in which the resulting particle positions are validated through the matching algorithm. The comparison with tomographic particleimage velocimetry data shows good agreement. The proposed algorithm allows this approach to be applied to volumetric measurements with seeding densities exceeding standard particle-tracking applications. Therefore, it can serve as a drop-in replacement for tomographic particle-image velocimetry at significantly reduced computational cost.

SerialTrack: ScalE and rotation invariant augmented Lagrangian particle tracking

“…Various algorithms have been created to detect and track individual particles [1] such as the straightforward k-nearestneighbor (kNN) searches, topology-based approaches where neighboring particles are employed to construct local surrounding topology features [27,36,[50][51][52], globally optimized search problems -including linear assignment programming [26], Kalman filtering [53], relaxation methods [19,54], and feature vectorbased techniques [22,25] (see a brief summary of particle tracking open-source codes in Table 2). Among these methods, the nearest neighbor-type search algorithms are typically suitable for relatively low numbers of particles that undergo displacements smaller than the typical interparticle separation distance.…”

SerialTrack: ScalE and Rotation Invariant Augmented Lagrangian Particle Tracking

“…Various algorithms have been created to detect and track individual particles [1] such as the straightforward k-nearest-neighbor (kNN) searches, topologybased approaches where neighboring particles are employed to construct local surrounding topology features [27,36,37,38,39], globally optimized search problems -including linear assignment programming [26], Kalman filtering [40], relaxation methods [19,41], and feature vector-based techniques [22,25] (see a brief summary of particle tracking open-source codes in Table 3). Among these methods, the nearest neighbor-type search algorithms are typically suitable for relatively low numbers of particles that undergo displacements smaller than the typical interparticle separation distance.…”

Image-based System for 3D Visualization of Flow in Hydrodynamic Tunnel

“…Then a set of synchronized high-speed cameras acquire video sequences of flow motion for further processing. The acquired registrations are processed in two separate steps, utilizing either Eulerian or Lagrangian approach [12,13,14,15]. Eulerian methods perform a voxel-based reconstruction of particles per time step, followed by 3D motion estimation, with some form of dense matching between the precomputed voxel grids from different time steps [16,12].…”