Live Webcast 15th Annual Charm++ Workshop

Particle tracking methods are widely used in fluid mechanics and multi-target tracking research because of their unique ability to reconstruct long trajectories with high spatial and temporal resolution. Recently, real-time image processing and particle localization has been accomplished through the use of specialized “smart cameras” that reduce data transfer rates up to 1000 times, extending measurement durations from seconds to weeks. Researchers have demonstrated 3D tracking of several objects in real-time, but as the number of objects is increased real-time tracking becomes impossible due to data transfer and processing bottlenecks. This problem may be solved by using parallel processing. However, very little has been published on parallel particle tracking algorithms that can fully utilize mulit-core processors and high performance computing clusters. In this paper, it is shown that parallelization of the particle tracking algorithm is a key step in achieving a scalable Lagrangian measurement system for particle tracking velocimetry (PTV) and may lead to real-time measurement capabilities. A parallel processing framework was developed based on frame decomposition and is programmed using the asynchronous object-oriented Charm++ paradigm. The parallel tracking algorithm was evaluated with three data sets including the PIV standard 3D images data set #352, a uniform data set for optimal parallel performance and a CFD generated non- uniform data set to test trajectory reconstruction accuracy, consistency with the sequential version, and scalability up to 512 processors. Ultimately, up to a 200-fold speedup is observed compared to the serial algorithm.