• 北大核心期刊(《中文核心期刊要目总览》2017版)
  • 中国科技核心期刊(中国科技论文统计源期刊)
  • JST 日本科学技术振兴机构数据库(日)收录期刊

留言板

尊敬的读者、作者、审稿人, 关于本刊的投稿、审稿、编辑和出版的任何问题, 您可以本页添加留言。我们将尽快给您答复。谢谢您的支持!

姓名
邮箱
手机号码
标题
留言内容
验证码

基于MDFPF-ResNet的红外手掌图像关键点定位

何小平 潘晴 田妮莉

何小平, 潘晴, 田妮莉. 基于MDFPF-ResNet的红外手掌图像关键点定位[J]. 微电子学与计算机, 2021, 38(10): 9-14. doi: 10.19304/J.ISSN1000-7180.2021.0135
引用本文: 何小平, 潘晴, 田妮莉. 基于MDFPF-ResNet的红外手掌图像关键点定位[J]. 微电子学与计算机, 2021, 38(10): 9-14. doi: 10.19304/J.ISSN1000-7180.2021.0135
HE Xiaoping, PAN Qing, TIAN Nili. Location of key points in infrared palm image based on MDFPF-ResNet[J]. Microelectronics & Computer, 2021, 38(10): 9-14. doi: 10.19304/J.ISSN1000-7180.2021.0135
Citation: HE Xiaoping, PAN Qing, TIAN Nili. Location of key points in infrared palm image based on MDFPF-ResNet[J]. Microelectronics & Computer, 2021, 38(10): 9-14. doi: 10.19304/J.ISSN1000-7180.2021.0135

基于MDFPF-ResNet的红外手掌图像关键点定位

doi: 10.19304/J.ISSN1000-7180.2021.0135
基金项目: 

国家自然科学基金 61901123

详细信息
    作者简介:

    何小平   男,(1994-),硕士研究生.研究方向为图像处理、图像识别. E-mail:1964109305@qq.com

    潘晴    男,(1975-),副教授,硕士研究生导师.研究方向为信号处理、模式识别、深度学习

    田妮莉   女,(1982-),博士,讲师.研究方向为图像及视频处理、时频分析

  • 中图分类号: TP391.4

Location of key points in infrared palm image based on MDFPF-ResNet

  • 摘要: 手诊作为一种能够推理人体器官健康状况的辅助诊断方法,在中医和智能中医的发展中占据着重要的地位.针对现有的红外图像关键点定位方法存在定位不准确的问题,提出一种多尺度空洞卷积特征金字塔融合的残差网络应用于红外手掌图像关键点定位的方法.在每个特征金字塔融合模块中,采用了空洞卷积的金字塔层级与改进型Bottleneck模块的并联结构,在增大感受野的同时,提升了残差网络的泛化能力;并将多个多尺度空洞卷积特征金字塔融合模块级联以逐步获取高层语义特征,在输出端加入全连接层和Dropout层得到关键点定位坐标.通过实验表明,所提方法在红外手掌九宫格图像上具有更好的定位性能,并具有较强的鲁棒性.
  • 图  1  特征提取模块

    图  2  Bottleneck模块及改进型

    图  3  MDFP层级

    图  4  MDFPF-ResNet网络结构

    图  5  手掌关键点位置分布图

    图  6  不同网络的关键点定位结果图对比

    图  7  不同网络的训练过程对比

    表  1  不同网络的关键点定位准确率对比

    网络 Key point A Key point B Key point C Key point D Key point E Key point F Key point G Key point H Key point I Average
    Open pose[6] 75.5 74.1 78.4 74.8 74.3 77.6 78.2 76.1 75.9 76.1
    HHLN[8] 83.2 84.8 80.4 81.9 79.5 84.9 83.3 82.1 80.4 82.8
    ResNet-101[11] 89.5 89.8 89.5 85.44 88.2 88.2 88.7 89.6 86.3 88.36
    文献[12] 93.2 90.1 93.3 88.2 90.5 91.2 92.5 89.6 90.7 91.03
    本文网络 94.4 93.6 91.2 92.5 96.6 95.3 94.8 95.8 93.8 94.2
    下载: 导出CSV

    表  2  不同网络的关键点定位召回率与测试时间参数大小的对比

    网络 Key point A Key point B Key point C Key point D Key point E Key point F Key point G Key point H Key point I Average 测试时间(ms) 网络参数(M)
    Open pose[6] 72.9 71.5 75.1 73.6 71.2 76.4 75.1 73.5 74.1 73.7 33 32.1
    HHLN[8] 83.4 81.5 83.2 80.2 81.1 80.8 82.6 81.9 81.1 81.7 42 28.6
    ResNet-101[11] 88.5 87.5 88.8 84.1 86.2 87.1 86.9 88.4 85.2 86.9 50 42.3
    文献[12] 92.3 89.4 92.6 87.4 89.8 90.4 91.8 89.1 90.1 90.3 60 35.2
    本文网络 93.1 92.3 89.5 91.6 95.3 94.8 94.2 95.1 92.9 93.2 58 38.4
    下载: 导出CSV
  • [1] 郭文静, 牛文民, 殷克敬. 殷克敬教授运用九宫八卦手诊法诊断心血管疾病经验[J]. 现代中医药, 2016, 36(5): 62-63. DOI:  10.13424/j.cnki.mtcm.2016.05.025.

    GUO W J, NIU W M, YIN K J. Professor Yin Kejing's experience in diagnosing cardiovascular diseases using the nine-gong and eight-diagram manual method[J]. Modern Traditional Chinese Medicine, 2016, 36(5): 62-63. DOI:  10.13424/j.cnki.mtcm.2016.05.025.
    [2] 李婷婷, 魏明, 李洪娟. 红外热像在中医学中的应用现状与展望[J]. 北京中医药大学学报(中医临床版), 2013, 20(4): 59-61. DOI:  10.3969/j.issn.1672-2205.2013.04.020.

    LI T T, WEI M, LI H J. The application status and prospect of infrared thermal imaging in traditional Chinese medicine[J]. Journal of Beijing University of Traditional Chinese Medicine (Clinical Medicine), 2013, 20(4): 59-61. DOI:  10.3969/j.issn.1672-2205.2013.04.020.
    [3] ZHAO Y F, ZHANG D, WANG Y X. Automatic location of facial acupuncture-point based on content of infrared thermal image[C]//Proceedings of the 5th International Conference on Computer Science & Education. Hefei, China: IEEE, 2010: 65-68. DOI: 10.1109/ICCSE.2010.5593668.
    [4] NEWELL A, YANG K Y, DENG J. Stacked hourglass networks for human pose estimation[C]//Proceedings of the 14th European Conference on Computer Vision. The Netherlands: Springer, 2016: 483-499. DOI: 10.1007/978-3-319-46484-8_29.
    [5] CHEN Y L, WANG Z C, PENG Y X, et al. Cascaded pyramid network for multi-person pose estimation[C]//Proceedings of 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Salt Lake City, UT, USA: IEEE, 2018: 7103-7112. DOI: 10.1109/CVPR.2018.00742.
    [6] CAO Z, SIMON T, WEI S E, et al. Realtime multi-person 2D pose estimation using part affinity fields[C]//Proceedings of 2017 IEEE Conference on Computer Vision and Pattern Recognition. Honolulu, HI, USA: IEEE, 2017: 1302-1310. DOI: 10.1109/CVPR.2017.143.
    [7] SUN K, XIAO B, LIU D, et al. Deep high-resolution representation learning for human pose estimation[C]//Proceedings of 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Long Beach, CA, USA: IEEE, 2019: 5693-5703. DOI: 10.1109/CVPR.2019.00584.
    [8] CHE Y L, SONG Y X, QI Y. A novel framework of hand localization and hand pose estimation[C]//Proceedings of 2019 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP). Brighton, UK: IEEE, 2019: 2222-2226. DOI: 10.1109/ICASSP.2019.8682382.
    [9] TANG W, WU Y. Does learning specific features for related parts help human pose estimation[C]//Proceedings of 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Long Beach, CA, USA: IEEE, 2019: 1107-1116. DOI: 10.1109/CVPR.2019.00120.
    [10] SUN X, XIAO B, WEI F Y, et al. Integral human pose regression[C]//Proceedings of the 15th European Conference on Computer Vision. Munich, Germany: Springer, 2018: 536-553. DOI: 10.1007/978-3-030-01231-1_33.
    [11] HE K M, ZHANG X Y, REN S Q, et al. Deep residual learning for image recognition[C]//Proceedings of 2016 IEEE Conference on Computer Vision and Pattern Recognition. Las Vegas, NV, USA: IEEE, 2016: 770-778. DOI: 10.1109/CVPR.2016.90.
    [12] TIANY, HU W, JIANG H S, et al. Densely connected attentional pyramid residual network for human pose estimation[J]. Neurocomputing, 2019, 347: 13-23. DOI:  10.1016/j.neucom.2019.01.104.
    [13] YAN C, WANG Y Q. A novel multi-user face detection under infrared illumination by real adaboost[C]//Proceedings of 2009 International Conference on Computational Intelligence and Software Engineering. Wuhan, China: IEEE, 2009: 1-6. DOI: 10.1109/CISE.2009.5366152.
    [14] HE K M, ZHANG X Y, REN S Q, et al. Identity mappings in deep residual networks[C]//Proceedings of the 14th European Conference on Computer Vision (ECCV). Amsterdam, The Netherlands: Springer, 2016: 630-645. DOI: 10.1007/978-3-319-46493-0_38.
    [15] XI R, HOU M S, FU M S, et al. Deep dilated convolution on multimodality time series for human activity recognition[C]//Proceedings of 2018 International Joint Conference on Neural Networks (IJCNN). Rio de Janeiro: IEEE, 2018: 1-8. DOI: 10.1109/IJCNN.2018.8489540.
  • 加载中
图(7) / 表(2)
计量
  • 文章访问数:  91
  • HTML全文浏览量:  47
  • PDF下载量:  26
  • 被引次数: 0
出版历程
  • 收稿日期:  2021-01-25
  • 修回日期:  2021-02-05
  • 刊出日期:  2021-10-05

目录

    /

    返回文章
    返回