Machine learning assisted droplet trajectories extraction in dense emulsions and their analysis

TL;DR

This paper combines machine learning algorithms with simulation data to analyze droplet trajectories in dense emulsions, revealing how local interactions influence droplet movement and informing dynamical modeling constraints.

Contribution

It introduces a novel application of YOLO and DeepSORT for trajectory extraction in dense emulsions and provides insights into droplet interaction dynamics.

Findings

Droplet movement is more affected by behind droplets than front droplets.

The analysis suggests constraints for modeling droplet dynamics in narrow channels.

Machine learning algorithms effectively extract trajectories in dense emulsion simulations.

Abstract

This work analyzes trajectories obtained by YOLO and DeepSORT algorithms of dense emulsion systems simulated by Lattice Boltzmann methods. The results indicate that the individual droplet's moving direction is influenced more by the droplets immediately behind it than the droplets in front of it. The analysis also provides hints on constraints on writing down a dynamical model of droplets for the dense emulsion in narrow channels.