中国科学院机构知识库网格
Chinese Academy of Sciences Institutional Repositories Grid
首页 机构 成果 学者
  
  1. CAS IR Grid
  2. 力学研究所
  3. 中国科学院力学研究所
  4. 流固耦合系统力学重点实验室(2012-)
A deep learning-based segmentation method for multi-scale and overlapped bubbles in gas-liquid bubbly flow

文献类型：期刊论文

作者Yang, Zhilong3; Tian, Wenbin3; Deng, Xiaoliang3; He, Xiashu3; Wang ZY(王志英)1; Wang JZ(王静竹)1,2; Wang YW(王一伟)1,2
刊名PHYSICS OF FLUIDS
出版日期2025-03-01
卷号37期号:3页码:17
ISSN号1070-6631
DOI10.1063/5.0254679
通讯作者Tian, Wenbin(wenbin.tian@cau.edu.cn) ; Wang, Zhiying(wangzhiying@imech.ac.cn)
英文摘要In the realm of fluid dynamics, gas-liquid bubbly flow represents a prevalent and significant multiphase flow phenomenon. With the advancement of imaging technology, high-speed photography combined with image processing techniques has become a common method for measuring bubbly flows. To overcome the challenges posed by multi-scale and overlapping bubbles in gas-liquid bubbly flows, a deep learning-based method for precise bubble contour segmentation and trajectory tracking has been developed. This approach involves specific optimizations and enhancements to the one-stage object detection model "You-Only-Look-Once version 8", leading to a bubble segmentation algorithm that strikes a balance between speed and precision. Omni-dimension dynamic convolution and high-resolution feature layer pyramid level 2 (P2) were integrated into the model to extract more precise spatial and texture information, enhancing precision and facilitating the detection of small-sized bubbles. Additionally, to address the issue of severe bubble overlap in images, the bubble spatially enhanced attention module was developed to capitalize on detailed texture, thereby achieving the segmentation of severely overlapping bubbles. Based on the improved detection model, combined with the Botsort tracking algorithm, vanishing bubble re-identification as well as continuous tracking of severely occluded bubbles are realized. The model achieves inference speeds of 0.427 s on central processing unit and 0.03 s on graphics processing unit (GPU), respectively, facilitating its application in efficiently processing large comprehensive datasets.
分类号一类/力学重要期刊
资助项目National Natural Science Foundation of China10.13039/501100001809[62371453] ; National Natural Science Foundation of China10.13039/501100001809[12372243] ; National Natural Science Foundation of China10.13039/501100001809[12293000] ; National Natural Science Foundation of China10.13039/501100001809[12293003] ; National Natural Science Foundation of China10.13039/501100001809[12293004] ; National Natural Science Foundation of China
WOS研究方向Mechanics ; Physics
语种英语
WOS记录号WOS:001447606800019
资助机构National Natural Science Foundation of China10.13039/501100001809 ; National Natural Science Foundation of China
其他责任者Tian, Wenbin，王志英
源URL[http://dspace.imech.ac.cn/handle/311007/100697]  
专题力学研究所_流固耦合系统力学重点实验室(2012-)
作者单位1.Chinese Acad Sci, Inst Mech, Beijing 100190, Peoples R China;
2.Univ Chinese Acad Sci, Sch Engn Sci, Beijing 100049, Peoples R China
3.China Agr Univ, Coll Engn, Beijing 100083, Peoples R China;
推荐引用方式
GB/T 7714		 Yang, Zhilong,Tian, Wenbin,Deng, Xiaoliang,et al. A deep learning-based segmentation method for multi-scale and overlapped bubbles in gas-liquid bubbly flow[J]. PHYSICS OF FLUIDS,2025,37(3):17.
APA Yang, Zhilong.,Tian, Wenbin.,Deng, Xiaoliang.,He, Xiashu.,王志英.,...&王一伟.(2025).A deep learning-based segmentation method for multi-scale and overlapped bubbles in gas-liquid bubbly flow.PHYSICS OF FLUIDS,37(3),17.
MLA Yang, Zhilong,et al."A deep learning-based segmentation method for multi-scale and overlapped bubbles in gas-liquid bubbly flow".PHYSICS OF FLUIDS 37.3(2025):17.

入库方式： OAI收割

来源：力学研究所

浏览0
下载0
收藏0
其他版本

除非特别说明，本系统中所有内容都受版权保护，并保留所有权利。