Deep learning-based automated and universal bubble detection and mask extraction in complex two-phase flows

Abstract While investigating multiphase flows experimentally, the spatiotemporal variation in the interfacial shape between different phases must be measured to analyze the transport phenomena. For this, numerous image processing techniques have been proposed, showing good performance. However, they...

Descripción completa

Guardado en:

Detalles Bibliográficos
Autores principales:	Yewon Kim, Hyungmin Park
Formato:	article
Lenguaje:	EN
Publicado:	Nature Portfolio 2021
Materias:	Medicine R Science Q
Acceso en línea:	https://doaj.org/article/b4c2274d734e4dd996b87f79e18e4921
Etiquetas:	Agregar Etiqueta Sin Etiquetas, Sea el primero en etiquetar este registro!

id	oai:doaj.org-article:b4c2274d734e4dd996b87f79e18e4921
record_format	dspace
spelling	oai:doaj.org-article:b4c2274d734e4dd996b87f79e18e49212021-12-02T17:39:31ZDeep learning-based automated and universal bubble detection and mask extraction in complex two-phase flows10.1038/s41598-021-88334-02045-2322https://doaj.org/article/b4c2274d734e4dd996b87f79e18e49212021-04-01T00:00:00Zhttps://doi.org/10.1038/s41598-021-88334-0https://doaj.org/toc/2045-2322Abstract While investigating multiphase flows experimentally, the spatiotemporal variation in the interfacial shape between different phases must be measured to analyze the transport phenomena. For this, numerous image processing techniques have been proposed, showing good performance. However, they require trial-and-error optimization of thresholding parameters, which are not universal for all experimental conditions; thus, their accuracy is highly dependent on human experience, and the overall processing cost is high. Motivated by the remarkable improvements in deep learning-based image processing, we trained the Mask R-CNN to develop an automated bubble detection and mask extraction tool that works universally in gas–liquid two-phase flows. The training dataset was rigorously optimized to improve the model performance and delay overfitting with a finite amount of data. The range of detectable bubble size (particularly smaller bubbles) could be extended using a customized weighted loss function. Validation with different bubbly flows yields promising results, with AP50 reaching 98%. Even while testing with bubble-swarm flows not included in the training set, the model detects more than 95% of the bubbles, which is equivalent or superior to conventional image processing methods. The pure processing speed for mask extraction is more than twice as fast as conventional approaches, even without counting the time required for tedious threshold parameter tuning. The present bubble detection and mask extraction tool is available online ( https://github.com/ywflow/BubMask ).Yewon KimHyungmin ParkNature PortfolioarticleMedicineRScienceQENScientific Reports, Vol 11, Iss 1, Pp 1-11 (2021)
institution	DOAJ
collection	DOAJ
language	EN
topic	Medicine R Science Q
spellingShingle	Medicine R Science Q Yewon Kim Hyungmin Park Deep learning-based automated and universal bubble detection and mask extraction in complex two-phase flows
description	Abstract While investigating multiphase flows experimentally, the spatiotemporal variation in the interfacial shape between different phases must be measured to analyze the transport phenomena. For this, numerous image processing techniques have been proposed, showing good performance. However, they require trial-and-error optimization of thresholding parameters, which are not universal for all experimental conditions; thus, their accuracy is highly dependent on human experience, and the overall processing cost is high. Motivated by the remarkable improvements in deep learning-based image processing, we trained the Mask R-CNN to develop an automated bubble detection and mask extraction tool that works universally in gas–liquid two-phase flows. The training dataset was rigorously optimized to improve the model performance and delay overfitting with a finite amount of data. The range of detectable bubble size (particularly smaller bubbles) could be extended using a customized weighted loss function. Validation with different bubbly flows yields promising results, with AP50 reaching 98%. Even while testing with bubble-swarm flows not included in the training set, the model detects more than 95% of the bubbles, which is equivalent or superior to conventional image processing methods. The pure processing speed for mask extraction is more than twice as fast as conventional approaches, even without counting the time required for tedious threshold parameter tuning. The present bubble detection and mask extraction tool is available online ( https://github.com/ywflow/BubMask ).
format	article
author	Yewon Kim Hyungmin Park
author_facet	Yewon Kim Hyungmin Park
author_sort	Yewon Kim
title	Deep learning-based automated and universal bubble detection and mask extraction in complex two-phase flows
title_short	Deep learning-based automated and universal bubble detection and mask extraction in complex two-phase flows
title_full	Deep learning-based automated and universal bubble detection and mask extraction in complex two-phase flows
title_fullStr	Deep learning-based automated and universal bubble detection and mask extraction in complex two-phase flows
title_full_unstemmed	Deep learning-based automated and universal bubble detection and mask extraction in complex two-phase flows
title_sort	deep learning-based automated and universal bubble detection and mask extraction in complex two-phase flows
publisher	Nature Portfolio
publishDate	2021
url	https://doaj.org/article/b4c2274d734e4dd996b87f79e18e4921
work_keys_str_mv	AT yewonkim deeplearningbasedautomatedanduniversalbubbledetectionandmaskextractionincomplextwophaseflows AT hyungminpark deeplearningbasedautomatedanduniversalbubbledetectionandmaskextractionincomplextwophaseflows
_version_	1718379843073081344

Deep learning-based automated and universal bubble detection and mask extraction in complex two-phase flows

Ejemplares similares