基于无载波超宽带雷达的小样本人体动作识别

doi:10.3969/j.issn.0372-2112.2020.03.025

电子学报 ›› 2020, Vol. 48 ›› Issue (3): 602-615.DOI: 10.3969/j.issn.0372-2112.2020.03.025

基于无载波超宽带雷达的小样本人体动作识别

蒋留兵^1,3, 周小龙^2,3, 车俐^1,3

1. 桂林电子科技大学计算机与信息安全学院, 广西桂林 541004;
2. 桂林电子科技大学信息与通信学院, 广西桂林 541004;
3. 桂林电子科技大学广西无线宽带通信与信号处理重点实验室, 广西桂林 541004

收稿日期:2018-11-14 修回日期:2019-07-21 出版日期:2020-03-25
- 作者简介:
- 蒋留兵 男,1973出生,江苏泰兴人,1997年于电子科技大学获得学士学位,2006年于南京理工大学获得硕士学位,1997年至2007年在中国电子科技集团公司第十四研究所从事雷达系统的研制工作.现为桂林电子科技大学研究员,主要方向为宽带信号与信息处理、智能信息处理和人工智能在雷达系统中的应用.E-mail:jlbnj@163.com;周小龙 男,1991年出生,江西九江人,现为桂林电子科技大学硕士研究生,主要研究方向为雷达信号处理、超宽带雷达人体动作检测与识别和人工智能在雷达系统中的应用.E-mail:jeff_zhou19910914@163.com
- 基金资助:
- 国家自然科学基金项目 (No.61561010）; 广西自然科学基金项目 (No.2017GXNSFAA198089）; 广西重点研发计划项目 (No.桂科AB18126003，No.桂科AB18221016）

Few-Shot Learning for Human Motion Recognition Based on Carrier-Free UWB Radar

JIANG Liu-bing^1,3, ZHOU Xiao-long^2,3, CHE Li^1,3

1. School of Computer and Information Security, Guilin University of Electronic Technology, Guilin, Guangxi 541004, China;
2. School of Information and Communication, Guilin University of Electronic Technology, Guilin, Guangxi 541004, China;
3. Key Laboratory of Wireless Broadband Communication and Signal Processing in Guangxi, Guilin University of Electronic Technology, Guilin, Guangxi 541004, China

Received:2018-11-14 Revised:2019-07-21 Online:2020-03-25 Published:2020-03-25

摘要/Abstract

摘要： 随着雷达硬件平台尺寸越来越小、成本越来越低，室内基于雷达的人体动作识别应用已经成为现实，能够在具有简单架构的低成本设备中实现.无载波超宽带雷达具有极高的分辨力，能够捕获人体细微动作变化并且对室内复杂环境具有很强的抗干扰能力.与基于视频人体动作识别研究相比，超宽带雷达还具有穿透家具、墙体以及保护个人隐私等优点.针对雷达回波信号利用传统时频分析方法实现人体动作识别比较耗时、实时性不好的缺陷，引入机器学习方法对不同类型人体动作进行分类识别.引入机器学习方法用于超宽带雷达人体动作识别最大难点是只有少量可用的超宽带雷达实测数据样本，针对该问题提出基于主成分分析法（PCA）和离散余弦变换（DCT）相结合的人体动作特征提取方法，并利用改进网格搜索算法优化的支持向量机在小样本数据下对人体动作进行识别，最后根据实测数据采取三种不同方案进行仿真实验，结果表明即使在训练数据样本只有5组的条件下，基于PCA和DCT相结合特征提取方法在不同类型人体动作的平均识别率均能达到96%以上.

关键词: 无载波超宽带雷达, 人体动作识别, 主成分分析法, 离散余弦变换, 小样本学习, 机器学习

Abstract: As radar hardware platforms become smaller and cheaper,indoor radar-based motion recognition applications have become reality and can be implemented in low-cost devices with simple architectures.The carrier-free ultra-wideband (UWB) radar has extremely high resolution,which can capture the slight movement of the human motion and has a strong anti-jamming capability in indoor complex environments.Human motion recognition based on UWB radar compared to video-based also has the advantage of penetrating furniture,walls and protecting personal privacy.Aiming at the defects that the traditional time-frequency analysis method based on radar realizes the human motion recognition is time-consuming and poor real-time performance,the machine learning method is introduced to classify and recognize different types of human motions.The biggest difficulty in introducing machine learning methods for UWB radar human motion recognition is that there are only a few-shot of available UWB radar measured data samples.Therefore,a human motion feature extraction method based on principal component analysis (PCA) and discrete cosine transform (DCT) is proposed.And the support vector machine (SVM) optimized by the improved grid search algorithm is used for human motion recognition under few-shot samples.Finally,simulations experiments are performed based on measured data through three different schemes.Under the condition that there are only 5 groups of training data samples,the average recognition rate of human motion recognition can reach more than 96%.

Key words: carrier-free UWB radar, human action recognition, principal component analysis (PCA), discrete cosine transform (DCT), few-shot learning, machine learning

中图分类号:

TP391.4

蒋留兵, 周小龙, 车俐. 基于无载波超宽带雷达的小样本人体动作识别[J]. 电子学报, 2020, 48(3): 602-615.

JIANG Liu-bing, ZHOU Xiao-long, CHE Li. Few-Shot Learning for Human Motion Recognition Based on Carrier-Free UWB Radar[J]. Acta Electronica Sinica, 2020, 48(3): 602-615.

参考文献

[1] Seyfio?lu M S,Özbayo?lu A M,Gürbüz S Z.Deep convolutional autoencoder for radar-based classification of similar aided and unaided human activities[J].IEEE Transactions on Aerospace and Electronic Systems,2018,54(4):1709-1723.
[2] Ding C,Zhang L,Gu C,et al.Non-contact human motion recognition based on UWB radar[J].IEEE Journal on Emerging and Selected Topics in Circuits and Systems,2018,8(2):306-315.
[3] Fioranelli F,Ritchie M,Griffiths H.Analysis of polarimetric multistatic human micro-Doppler classification of armed/unarmed personnel[A].2015 Radar Conference (RadarCon)[C].Arlington,VA,USA:IEEE,2015.0432-0437.
[4] Giansanti D,Maccioni G,Macellari V.The development and test of a device for the reconstruction of 3-D position and orientation by means of a kinematic sensor assembly with rate gyroscopes and accelerometers[J].IEEE Transactions on Biomedical Engineering,2005,52(7):1271-1277.
[5] Anderson D,Luke R H,Keller J M,et al.Linguisticsummarization of video for fall detection using voxel person and fuzzy logic[J].Computer Vision & Image Understanding Cviu,2009,113(1):80-89.
[6] Sixsmith A,Johnson N,Whatmore R.Pyroelectric IR sensor arrays for fall detection in the older population[A].Journal de Physique IV (Proceedings)[C].France:EDP Sciences,2005,128(1):153-160.
[7] Yazar A,Keskin F,Töreyin B U,et al.Fall detection using single-tree complex wavelet transform[J].Pattern Recognition Letters,2013,34(15):1945-1952.
[8] Augusto J C,Nakashima H,Aghajan H.Ambient Intelligence and Smart Environments:A State of the Art[M].Boston,MA:Springer, 2010.3-31.
[9] Augusto J C,McCullagh P.Ambient intelligence:Concepts and applications[J].Computer Science and Information Systems,2007,4(1):1-27.
[10] Ramos C,Augusto J C,Shapiro D.Ambient intelligence-the next step for artificialintelligence[J].IEEE Intelligent Systems,2008,23(2):15-18.
[11] Liu L,Popescu M,Skubic M,et al.An automatic fall detection framework using data fusion of Doppler radar and motion sensor network[A].201436th Annual International Conference of the IEEE,Engineering in Medicine and Biology Society (EMBC)[C].Chicago,Illinois,USA:IEEE,2014.5940-5943.
[12] Li C,Lubecke V M,Boric-Lubecke O,et al.A review on recent advances in Doppler radar sensors for noncontact healthcare monitoring[J].IEEE Transactions on Microwave Theory and Techniques,2013,61(5):2046-2060.
[13] Lien J,Gillian N,Karagozler M E,et al.Soli:Ubiquitous gesture sensing with millimeter wave radar[J].ACM Transactions on Graphics (TOG),2016,35(4):142.
[14] Fan T,Ma C,Gu Z,et al.Wireless hand gesture recognition based on continuous-wave Doppler radar sensors[J].IEEE Transactions on Microwave Theory and Techniques,2016,64(11):4012-4020.
[15] Fioranelli F,Ritchie M,Griffiths H.Multistatic human micro-Doppler classification of armed/unarmed personnel[J].IET Radar,Sonar & Navigation,2015,9(7):857-865.
[16] Yarovoy A G,Zhuge X,Savelyev T G,et al.Comparison of UWB technologies for human being detection with radar[A].2007 European Radar Conference[C].Munich,Germany:IEEE,2007.1574-1577.
[17] Liu L,Liu S.Remote detection of human vital sign with stepped-frequency continuous wave radar[J].IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing,2014,7(3):775-782.
[18] Lyonnet B,Ioana C,Amin M G.Human gait classification using microdoppler time-frequency signal representations[A].2010 IEEE Radar Conference[C].Washington,DC,USA:IEEE,2010.915-919.
[19] Kim Y,Ling H.Human activity classification based on micro-Doppler signatures using a support vectormachine[J].IEEE Transactions on Geoscience and Remote Sensing,2009,47(5):1328-1337.
[20] Zenaldin M,Narayanan R M.Features associated with radar micro-Doppler signatures of various human activities[A].SPIE Conference on Radar Sensor Technology XIX and Active and Passive Signatures VI[C].Baltimore,Maryland,United States:SPIE,2015.94611D-1-94611D-14.
[21] Lv H,Lu G H,Jing X J,et al.A new ultra-wideband radar for detecting survivors buried under earthquake rubbles[J].Microwave and Optical Technology Letters,2010,52(11):2621-2624.
[22] Molchanov P,Astola J,Egiazarian K,et al.Ground moving target classification by using DCT coefficients extracted from micro-Doppler radar signatures and artificial neuron network[A].2011 Microwaves,Radar and Remote Sensing Symposium[C].Kiev,Ukraine:IEEE,2011.173-176.
[23] 廖志成.基于超宽带雷达的人体姿态识别技术研究[D].南京:南京理工大学,2017. Liao Zhicheng.Research on Human Activity Recognition Technology Based on Ultra-wideband Radar[D].Nanjing:Nanjing University of Science & Technology,2017.(in Chinese)
[24] Bryan J D,Kwon J,Lee N,et al.Application of ultra-wide band radar for classification of humanactivities[J].IET Radar,Sonar & Navigation,2012,6(3):172-179.
[25] Li J,Phung S L,Tivive F H C,et al.Automatic classification of human motions using Doppler radar[A].The 2012 International Joint Conference on Neural Networks (IJCNN)[C].Brisbane,QLD,Australia:IEEE,2012.1-6.
[26] 陈旻罡.基于多普勒雷达的跌倒检测方法研究[D].太原:太原理工大学,2018. Chen Mingang.Fall Detection Based on Doppler Radar[D].Taiyuan:Taiyuan University of Technology,2018.(in chinese)
[27] 李康顺,韦蕴珊,张文生.小生境演化算法下的WDCT图像压缩方法[J].电子学报,2014,42(4):809-814. Li Kangshun,Wei Yunshan,Zhang Wensheng.WDCT image compresion based on niching evolutionary algorithm[J].Acta Electronica Sinica,2014,4(42):809-814.(in Chinese)
[28] 万飞.基于网格搜索的支持向量机在入侵检测中的应用[D].合肥:合肥工业大学,2015. Wan Fei.The Application of Support Vector Machine Based on Grid Search in Network Intrusion Detection[D].Heifei:Heifei University of Technology,2015.(in Chinese)
[29] 温博文,董文瀚,解武杰,等.基于改进网格搜索算法的随机森林参数优化[J].计算机工程与应用,2018,54:(10):154-157. Wen Bowen,Dong Wenhan,Xie Wujie,et al.Parameter optimization method for random forest based on improved grid search algorithm[J].CEA,2018,54(10):154-157.(in Chinese)
[30] 戴舜,朱方,徐艳云,等.基于PCA与EMD的超宽带雷达生命信号检测算法[J].电子学报,2012,40(2):344-349. Dai Shun,Zhu Fang,Xu Yanyun,et al.Vital signal detection method based on principal component analysis and empirical mode decomposition for ultra-wideband radar[J].Acta Electronica Sinica,2012,40(2):344-349.(in Chinese)
[31] 高尚,杨静宇.群智能算法及其应用[M].北京:中国水利水电出版社,2006. Gao Shang,Yang Jinyu.Group Intelligence Algorithm and Its Application[M].Beijing:China Water Conservancy and Hydropower Press,2006.(in Chinese)
[32] 罗博文.基于支持向量机的人体动作识别智能算法研究与实现[D].哈尔滨:哈尔滨工业大学,2016. Luo Bowen.Intelligent Human Action Recognition Algorithm Based on Support Vector Machine[D].Harbin:Harbin Insitute of Technology,2016.(in Chinese)
[33] Javier R J,Kim Y.Application of linear predictive coding for human activity classification based on micro-Dopplersignatures[J].IEEE Geoscience and Remote Sensing Letters,2014,11(10):1831-1834.
[34] Shao Y,Guo S,Sun L,et al.Human motion classification based on range information with deep convolutional neural network[A].20174th International Conference on Information Science and Control Engineering (ICISCE)[C].Changsha,China:IEEE,2017.1519-1523.
[35] WANG Jun,ZHU He,LEI Peng,et al.CNN based classification of rigid targets in space using radar micro-doppler signatures[J].Chinese Journal of Electronics,2019,28(4):856-862.

基于无载波超宽带雷达的小样本人体动作识别

Few-Shot Learning for Human Motion Recognition Based on Carrier-Free UWB Radar

PDF

可视化

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 15

编辑推荐

Metrics

[1]	李雪莹, 王田路, 梁鹏, 王翀. 基于系统模型的用户评论中非功能需求的自动分类[J]. 电子学报, 2022, 50(9): 2079-2089.
[2]	黄璐, 蔚保国, 李宏生, 李隽, 贾浩男, 程建强, 李雅宁. GNSS拒止环境下的伪卫星指纹定位方法[J]. 电子学报, 2022, 50(4): 811-822.
[3]	崔亚奇, 何友, 唐田田, 熊伟. 一种深度学习航迹关联方法[J]. 电子学报, 2022, 50(3): 759-763.
[4]	蒋华伟, 郭陶, 杨震. 车辆路径问题研究进展[J]. 电子学报, 2022, 50(2): 480-492.
[5]	郑锐, 汪秋云, 林卓庞, 靖蓉琦, 姜政伟, 傅建明, 汪姝玮. 一种基于威胁情报层次特征集成的挖矿恶意软件检测方法[J]. 电子学报, 2022, 50(11): 2707-2715.
[6]	张梦, 郑建宏, 刘香燕, 何云. 基于集成学习的全双工中继系统安全中继选择方案研究[J]. 电子学报, 2021, 49(9): 1852-1856.
[7]	郑海颖, 王峰, 姜维, 王志强, 姚西文. 神经网络训练策略对高分辨率遥感图像场景分类性能影响的评估[J]. 电子学报, 2021, 49(8): 1599-1614.
[8]	周东明, 张灿龙, 李志欣, 王智文. 基于多层级视觉融合的图像描述模型[J]. 电子学报, 2021, 49(7): 1286-1290.
[9]	侯秀聪, 赖晓平, 曹九稳. 正则化超限学习机的最大分划广义交替方向乘子法[J]. 电子学报, 2021, 49(4): 625-630.
[10]	陈长青, 郭春, 崔允贺, 申国伟, 蒋朝惠. 基于API短序列的勒索软件早期检测方法[J]. 电子学报, 2021, 49(3): 586-595.
[11]	刘润滋, 吴伟华, 张文柱, 周笛, 张琰. 基于图学习的密集空间网络传输资源调度方法[J]. 电子学报, 2021, 49(11): 2133-2137.
[12]	仇祝令, 查宇飞, 吴敏, 王青. 基于注意力学习的正则化相关滤波跟踪算法[J]. 电子学报, 2020, 48(9): 1762-1768.
[13]	宋宇琦, 高旻, 李骏东, 荣文戈, 熊庆宇. 网络欺凌检测综述[J]. 电子学报, 2020, 48(6): 1220-1229.
[14]	高鹏, 陈智华. 一种基于拓扑信息的预测疾病相关的MicroRNAs方法[J]. 电子学报, 2020, 48(2): 333-340.
[15]	王进, 高选人, 张睿, 孙开伟, 邓欣. 结合目标特定特征和目标相关性的多目标回归[J]. 电子学报, 2020, 48(11): 2092-2100.