基于共现流增强双向金字塔卷积网络的密集液滴识别

doi:10.19678/j.issn.1000-3428.0062060

计算机工程 ›› 2022, Vol. 48 ›› Issue (7): 241-246,253. doi: 10.19678/j.issn.1000-3428.0062060

基于共现流增强双向金字塔卷积网络的密集液滴识别

朱凌^1,2, 王雅萍¹, 廖丽敏²

1. 郑州大学信息工程学院, 郑州 450001;
2. 广东顺德创新设计研究院, 广东佛山 528311

收稿日期:2021-07-13 修回日期:2021-09-07 出版日期:2022-07-15 发布日期:2021-09-29
作者简介:朱凌(1995—),男,硕士研究生,主研方向为图像处理、目标检测;王雅萍,副教授;廖丽敏(通信作者),工程师。
基金资助:
国家自然科学基金（61772475）；国家自然科学青年基金（61906172）；河南省高等学校重点科研项目计划（20A510009）。

Dense Droplet Identification Based on Co-occurrence Flow Enhanced Bidirectional Pyramidal Convolution Network

ZHU Ling^1,2, WANG Yaping¹, LIAO Limin²

1. School of Information Engineering, Zhengzhou University, Zhengzhou 450001, China;
2. Guangdong Shunde Innovative Design Institute, Foshan, Guangdong 528311, China

Received:2021-07-13 Revised:2021-09-07 Online:2022-07-15 Published:2021-09-29

摘要/Abstract

摘要： 基于深度学习的数字聚合酶链式反应（PCR）液滴识别对PCR图像中的目标进行高阶语义建模，能够减少人工参与特征设计和筛选带来的误差，但忽略了目标的低层物理结构和几何外观细节信息，且在特征建模的过程中重复使用了大量冗余信息，对特征的表征能力有待改善。提出一种共现流增强双向金字塔卷积网络（CoF-BiPCN）框架用于PCR液滴识别和统计。为增强金字塔的内部和层间相关性，设计具有时空分支的双向金字塔卷积网络，从正反2个方向对金字塔卷积网络得到的多尺度特征进行聚合，模拟PCR图像中液滴的上下文语义以及不同层级的细节信息，以捕获液滴的物理外观等低层信息。同时，设计切片的共现注意力（SCo-AN）模块，将不同尺度的高低层信息在不同的切片子空间中进行共享聚合，并交叉传递到不同分支的BiPCN中，强化高低层特征信息的交互和依赖关系，进一步增强信息流对PCR图像上液滴的表征，实现低层和高阶信息流的共享与交叉聚合。实验结果表明，CoF-BiPCN具备良好的识别性能，准确率和平均精度均值分别达到84.74%和45.09%，与Cascade RCNN模型相比分别提高4.3和3.12个百分点。

关键词: 数字聚合酶链式反应液滴识别, 金字塔卷积网络, 多尺度信息, 共现注意力, 层间相关性, 交叉聚合

Abstract: A pyramid network faces a range of problems in recognizing digital Polymerase Chain Reaction(PCR) droplets, such as ignoring the physical appearance of droplets, and the internal structure and interlayer correlation of pyramids.To address these issues, this study proposes a co-occurrence flow enhanced Bidirectional Pyramid Convolution Network(BiPCN) framework for PCR droplet recognition and statistics framework.First, to enhance the internal and inter-layer correlation of the pyramid, a BiPCN with spatio-temporal branches is designed to capture the low-level information such as the physical appearance of droplets, so as to model the high-order semantics and context information of droplets.Second, Slice Co-occurrence Attention(SCo-AN) module is designed to further enhance the characterization of droplets on PCR images by information flow, so as to realize the sharing and cross polymerization of low-level and high-order information flow.The experimental results show that the proposed method has good recognition performance, that is, the accuracy and mean Average Precision(mAP) reach 84.74% and 45.09% respectively, which is more accurate than that of Cascade RCNN model, the accuracy and mAP are increased by 4.3 and 3.12 percentage points, respectively.

Key words: Polymerase Chain Reaction(PCR) droplet identification, Pyramid Convolutional Network(PCN), multi-scale information, Co-occurrence Attention(SCo-AN), inter-layer correlation, cross polymerization

中图分类号:

TP391

朱凌, 王雅萍, 廖丽敏. 基于共现流增强双向金字塔卷积网络的密集液滴识别[J]. 计算机工程, 2022, 48(7): 241-246,253.

ZHU Ling, WANG Yaping, LIAO Limin. Dense Droplet Identification Based on Co-occurrence Flow Enhanced Bidirectional Pyramidal Convolution Network[J]. Computer Engineering, 2022, 48(7): 241-246,253.

https://www.ecice06.com/CN/Y2022/V48/I7/241

图/表 7

20221029175324

20221029175328

20221029175332

20221029175336

20221029175340

20221029175343

20221029175347

参考文献

[1] 刘聪, 蒋克明, 周武平, 等.微滴技术的数字PCR研究现状及发展趋势[J].化学研究与应用, 2018, 30(7):1041-1047. LIU C, JIANG K M, ZHOU W P, et al.Research status and development trend of digital PCR with droplet technology[J].Chemical Research and Application, 2018, 30(7):1041-1047.(in Chinese)
[2] 邱梦凯.数字PCR液滴识别方法的研究与实现[D].沈阳:沈阳理工大学, 2019. QIU M K.Research and realization of digital PCR droplet recognition method[D].Shenyang:Shenyang Ligong University, 2019.(in Chinese)
[3] 刘松生, 袁浩均, 刘强, 等.液滴数字PCR芯片结果自动化读出平台的研究[J].现代电子技术, 2017, 40(18):1-6. LIU S S, YUAN H J, LIU Q, et al.Research on automatic result readout platform of droplet digital PCR chip[J].Modern Electronics Technique, 2017, 40(18):1-6.(in Chinese)
[4] 刘聪, 董文飞, 蒋克明, 等.基于改进分水岭分割算法的致密荧光微滴识别[J].中国光学, 2019, 12(4):783-790. LIU C, DONG W F, JIANG K M, et al.Recognition of dense fluorescent droplets using an improved watershed segmentation algorithm[J].Chinese Optics, 2019, 12(4):783-790.(in Chinese)
[5] MOROKOFF A, JONES J, NGUYEN H, et al.Serum microRNA is a biomarker for post-operative monitoring in glioma[J].Journal of Neuro-Oncology, 2020, 149(3):391-400.
[6] KASTRATI Z, DALIPI F, IMRAN A S, et al.Sentiment analysis of students' feedback with NLP and deep learning:a systematic mapping study[J].Applied Sciences, 2021, 11(9):1-23.
[7] PEKALA M, JOSHI N, LIU T Y A, et al.OCT segmentation via deep learning:a review of recent work[C]//Proceedings of Asian Conference on Computer Vision.Berlin, Germany:Springer, 2018:316-322.
[8] OH S, CHANG A J, ASHAPURE A, et al.Plant counting of cotton from UAS imagery using deep learning-based object detection framework[J].Remote Sensing, 2020, 12(18):2981.
[9] HU Z M, FANG W B, GOU T, et al.A novel method based on a Mask R-CNN model for processing dPCR images[J].Analytical Methods, 2019, 11(27):3410-3418.
[10] ÇALLı E, MURPHY K, KURSTJENS S, et al.Deep learning with robustness to missing data:a novel approach to the detection of COVID-19[J].PLoS One, 2021, 16(7):1-10.
[11] CHIU W H K, VARDHANABHUTI V, POPLAVSKIY D, et al.Detection of COVID-19 using deep learning algorithms on chest radiographs[J].Journal of Thoracic Imaging, 2020, 35(6):369-376.
[12] DURVE M, BONACCORSO F, MONTESSORI A, et al.A fast and efficient deep learning procedure for tracking droplet motion in dense microfluidic emulsions[J].Philosophical Transactions of the Royal Society A:Mathematical, Physical and Engineering Sciences, 2021, 379(2208):1-10.
[13] 邹贞贞.基于稀疏表示和深度学习的液滴图像超分辨重建[D].北京:中国科学院大学, 2018. ZOU Z Z.Droplet image super resolution based on sparse representation and deep learning[D].Beijing:University of Chinese Academy of Sciences, 2018.(in Chinese)
[14] HU S P, GAO Y, NIU Z M, et al.Weakly supervised deep learning for COVID-19 infection detection and classification from CT images[J].IEEE Access, 2020, 8:118869-118883.
[15] DUTA I C, LIU L, ZHU F, et al.Pyramidal convolution:rethinking convolutional neural networks for visual recognition[EB/OL].[2021-06-20].https://arxiv.org/abs/2006.11538.
[16] HSIEH T I, LO Y C, CHEN H T, et al.One-shot object detection with co-attention and co-excitation[EB/OL].[2021-06-20].https://arxiv.org/abs/1911.12529.
[17] ZHANG Q L, YANG Y B.SA-Net:shuffle attention for deep convolutional neural networks[C]//Proceedings of IEEE International Conference on Acoustics, Speech and Signal Processing.Washington D.C., USA:IEEE Press, 2021:2235-2239.
[18] HE K M, ZHANG X Y, REN S Q, et al.Deep residual learning for image recognition[C]//Proceedings of IEEE Conference on Computer Vision and Pattern Recognition.Washington D.C., USA:IEEE Press, 2016:770-778.
[19] LIN T Y, DOLLÁR P, GIRSHICK R, et al.Feature pyramid networks for object detection[C]//Proceedings of IEEE Conference on Computer Vision and Pattern Recognition.Washington D.C., USA:IEEE Press, 2017:936-944.
[20] REDMON J, FARHADI A.YOLOv3:an incremental improvement[EB/OL].[2021-06-20].https://arxiv.org/abs/1804.02767.
[21] REN S Q, HE K M, GIRSHICK R, et al.Faster R-CNN:towards real-time object detection with region proposal networks[J].IEEE Transactions on Pattern Analysis and Machine Intelligence, 2017, 39(6):1137-1149.
[22] HE K M, GKIOXARI G, DOLLÁR P, et al.Mask R-CNN[C]//Proceedings of IEEE International Conference on Computer Vision.Washington D.C., USA:IEEE Press, 2017:2980-2988.
[23] CAI Z W, VASCONCELOS N.Cascade R-CNN:delving into high quality object detection[C]//Proceedings of IEEE/CVF Conference on Computer Vision and Pattern Recognition.Washington D.C., USA:IEEE Press, 2018:6154-6162.

选择文件类型/文献管理软件名称

选择包含的内容

基于共现流增强双向金字塔卷积网络的密集液滴识别

Dense Droplet Identification Based on Co-occurrence Flow Enhanced Bidirectional Pyramidal Convolution Network

RichHTML

PDF

可视化

被引次数

摘要/Abstract

引用本文

使用本文

图/表 7

参考文献

相关文章 2

编辑推荐

Metrics

本文评价

[1]	张雯雯, 徐杨, 白芮, 陈娜. 基于改进堆叠沙漏网络的动物姿态估计[J]. 计算机工程, 2023, 49(2): 263-270.
[2]	张永梅,付昊天,孙海燕,张睿,陈立潮,潘理虎. 一种基于改进FCN的多光谱图像建筑物识别方法[J]. 计算机工程, 2019, 45(1): 239-245.

模态框（Modal）标题

选择文件类型/文献管理软件名称

选择包含的内容

基于共现流增强双向金字塔卷积网络的密集液滴识别

Dense Droplet Identification Based on Co-occurrence Flow Enhanced Bidirectional Pyramidal Convolution Network

RichHTML

PDF

可视化

被引次数

摘要/Abstract

引用本文

使用本文

图/表 7

参考文献

相关文章 2

编辑推荐

Metrics

本文评价