skip to main content
US FlagAn official website of the United States government
dot gov icon
Official websites use .gov
A .gov website belongs to an official government organization in the United States.
https lock icon
Secure .gov websites use HTTPS
A lock ( lock ) or https:// means you've safely connected to the .gov website. Share sensitive information only on official, secure websites.

Explore Research Products in the PAR
It may take a few hours for recently added research products to appear in PAR search results.


Title: DropletMask: Leveraging visual data for droplet impact analysis
Abstract Machine learning‐assisted computer vision represents a state‐of‐the‐art technique for extracting meaningful features from visual data autonomously. This approach facilitates the quantitative analysis of images, enabling object detection and tracking. In this study, we utilize advanced computer vision to precisely identify droplet motions and quantify their impact forces with spatiotemporal resolution at the picoliter or millisecond scale. Droplets, captured by a high‐speed camera, are denoised through neuromorphic image processing. These processed images are employed to train convolutional neural networks, allowing the creation of segmented masks and bounding boxes around moving droplets. The trained networks further digitize time‐varying multi‐dimensional droplet features, such as droplet diameters, spreading and sliding motions, and corresponding impact forces. Our innovative method offers accurate measurement of small impact forces with a resolution of approximately 10 pico‐newtons for droplets in the micrometer range across various configurations with the time resolution at hundreds of microseconds.  more » « less
Award ID(s):
2045322
PAR ID:
10633001
Author(s) / Creator(s):
; ;
Publisher / Repository:
Wiley
Date Published:
Journal Name:
Droplet
Volume:
3
Issue:
4
ISSN:
2769-2159
Subject(s) / Keyword(s):
droplet masks
Format(s):
Medium: X
Sponsoring Org:
National Science Foundation
More Like this
  1. ; ; ; ; ; (, Advanced Materials)
    Abstract A key challenge underlying the design of miniature machines is encoding materials with time‐ and space‐specific functional behaviors that require little human intervention. Dissipative processes that drive materials beyond equilibrium and evolve continuously with time and location represent one promising strategy to achieve such complex functions. This work reports how internal nonequilibrium states of liquid crystal (LC) emulsion droplets undergoing chemotaxis can be used to time the delivery of a chemical agent to a targeted location. During ballistic motion, hydrodynamic shear forces dominate LC elastic interactions, dispersing microdroplet inclusions (microcargo) within double emulsion droplets. Scale‐dependent colloidal forces then hinder the escape of dispersed microcargo from the propelling droplet. Upon arrival at the targeted location, a circulatory flow of diminished strength allows the microcargo to cluster within the LC elastic environment such that hydrodynamic forces grow to exceed colloidal forces and thus trigger the escape of the microcargo. This work illustrates the utility of the approach by using microcargo that initiate polymerization upon release through the outer interface of the carrier droplet. These findings provide a platform that utilizes nonequilibrium strategies to design autonomous spatial and temporal functions into active materials. 
    more » « less
  2. ; ; ; ; ; (, Springer Nature Switzerland)
    Cerebrospinal fluid (CSF) plays a critical role in brain metabolism and protection from external forces. Traditional MRI can provide some insights into CSF dynamics; however, more advanced and cost-effective methods are needed for precise and comprehensive visualization of flow patterns, velocities, and directions in clinical settings. In this paper, we demonstrate a new application of a few open-source computer vision software packages to capture CSF motion from time spatial inversion pulse (Time-SLIP) MRI clinical images (in DICOM format). To test the hypothesis that the CSF flow depends on head motions, a reliable and robust pipeline of processing Time-SLIP MRI images is developed to extract both anatomy and CSF motion dynamics. The paper presents a methodology for extracting unsteady flow information from Time-SLIP MRI images and the results of its application. The results show that the computer vision method can be applied to extract unsteady CSF flow information. We also discuss observations and identify future areas for improvement by integrating CFD simulations for validation as a vital component for studying CSF dynamics. 
    more » « less
  3. ; ; ; ; (, Physics of Fluids)
    Advection-enhanced heat and mass transport from a single droplet neutrally suspended in a simple shear flow has been studied using high-fidelity numerical simulation. The capillary number ranges from 0.01 to 0.5, which encompasses the entire range of small deformation, large deformation, and breakup of the droplets. The Reynolds number is from 0.01 to 1, including regions of both weak and strong advection. The temperature and mass concentration are modeled as the concentration of a passive scalar released at the droplet surface. Two Schmidt numbers, 10 and 100, are considered, for which flow advection plays a role in the transport of passive scalar. For unbroken droplets, the interaction between the carrier fluid and the suspended droplet leads to several different flows around the droplet. The fluid motions together with scalar diffusion constitute a coupled transport mechanism for passive scalar. The dependence of scalar release rate on Reynolds and Peclet numbers can be roughly described by the correlation for a rigid sphere. For broken droplets, the basic flow features around the droplet during the process of elongation and breakup are similar to those of an unbroken droplet. The variation of the scalar release rate can be decomposed into several stages, corresponding to the process of droplet elongation and breakup. The variation of the scalar release rate exhibits a high correlation with the capillary, Reynolds, and Peclet numbers. This suggests that it is feasible to develop an empirical model that incorporates the effects of the number and size distributions of child droplets after breakup. 
    more » « less
  4. ; ; ; ; ; ; ; ; (, npj Computational Materials)
    Abstract In computer vision, single-image super-resolution (SISR) has been extensively explored using convolutional neural networks (CNNs) on optical images, but images outside this domain, such as those from scientific experiments, are not well investigated. Experimental data is often gathered using non-optical methods, which alters the metrics for image quality. One such example is electron backscatter diffraction (EBSD), a materials characterization technique that maps crystal arrangement in solid materials, which provides insight into processing, structure, and property relationships. We present a broadly adaptable approach for applying state-of-art SISR networks to generate super-resolved EBSD orientation maps. This approach includes quaternion-based orientation recognition, loss functions that consider rotational effects and crystallographic symmetry, and an inference pipeline to convert network output into established visualization formats for EBSD maps. The ability to generate physically accurate, high-resolution EBSD maps with super-resolution enables high-throughput characterization and broadens the capture capabilities for three-dimensional experimental EBSD datasets. 
    more » « less
  5. ; ; ; ; (, Journal of Manufacturing Science and Engineering)
    Abstract Inkjet printing (IJP) is one of the promising additive manufacturing techniques that yield many innovations in electronic and biomedical products. In IJP, the products are fabricated by depositing droplets on substrates, and the quality of the products is highly affected by the droplet pinch-off behaviors. Therefore, identifying pinch-off behaviors of droplets is critical. However, annotating the pinch-off behaviors is burdensome since a large amount of images of pinch-off behaviors can be collected. Active learning (AL) is a machine learning technique which extracts human knowledge by iteratively acquiring human annotation and updating the classification model for the pinch-off behaviors identification. Consequently, a good classification performance can be achieved with limited labels. However, during the query process, the most informative instances (i.e., images) are varying and most query strategies in AL cannot handle these dynamics since they are handcrafted. Thus, this paper proposes a multiclass reinforced active learning (MCRAL) framework in which a query strategy is trained by reinforcement learning (RL). We designed a unique intrinsic reward signal to improve the classification model performance. Moreover, how to extract the features from images for pinch-off behavior identification is not trivial. Thus, we used a graph convolutional network for droplet image feature extraction. The results show that MCRAL excels AL and can reduce human efforts in pinch-off behavior identification. We further demonstrated that, by linking the process parameters to the predicted droplet pinch-off behaviors, the droplet pinch-off behavior can be adjusted based on MCRAL. 
    more » « less
  • Citation Formats
  • MLA
  • APA
  • Chicago
  • BibTeX
  • Text Encoded Conversion
  • XML
  • Save / Share this Record
  • Send to Email