一种基于无源射频技术的用户步态识别及认证方法

doi:10.16180/j.cnki.issn1007-7820.2020.06.001

摘要/Abstract

摘要：

步态识别作为一种新兴的生物特征识别技术,具有距离远、难伪造的优点,在智能监控等领域中具有广泛的应用前景。现有的步态识别方法存在着算法计算复杂、用户参与度高、设备开销较大等问题。针对这些问题,文中提出了一种基于无源射频技术的用户步态识别方法。该方法使用了门禁系统中已经广泛部署的无源射频标签,通过标签相位数据计算用户行走动态速度,利用多标签信号互补特性进行信号补偿,提取用户步态频率响应特征。实验结果显示,该方法的用户识别准确率高达91.87%,特征提取及比对的时延仅有0.129 s。在训练数据极少、用户参与度低的情况下,实现了高效率、较准确的用户步态识别及认证。

关键词: 无源射频技术, 步态识别, 用户行为感知, 移动计算, 无线感知, 物联网, 身份认证, 生物特征识别

Abstract:

As a rising biometric features recognition technology, human gait recognition has the advantages of long distance and difficult to be forged. As a result, human gait recognition has been widely used in many applications, including smart monitoring system. Current human gait recognition methods have their own problems of complex algorithms calculation, high user participation and extra hardware requirements. Aiming at these problems, a new method for user gait recognition based on passive radio frequency technology was proposed in this study. This method used the passive radio frequency tags that have been widely deployed in the access control system. The proposed method calculated the dynamic moving speed of the user by analyzing the phase profiles of tags, compensated the samples by considering multiple signal profiles of tags, and extracted the frequency domain recognition features. The experiments showed that this method could achieve an accuracy up to 91.87% , and the time delay was only 0.129 s. In conclusion, this method realized a highly efficient and accurate human gait recognition and identification with limited training data and user participation.

Key words: passive RFID, human gait recognition, human movement detection, mobile computing, wireless sensing, internet of things, user identification, biometric features recognition

中图分类号:

TP391

王鸽,惠维,丁菡,赵鲲,赵季中. 一种基于无源射频技术的用户步态识别及认证方法[J]. 电子科技, 2020, 33(6): 1-7.

WANG Ge,XI Wei,DING Han,ZHAO Kun,ZHAO Jizhong. Human Gait Recognition and Identification with Passive RFID[J]. Electronic Science and Technology, 2020, 33(6): 1-7.

图/表 10

图1

图2

图3

图4

图5

图6

表1

图7

表2

表3

参考文献 17

[1]	何逸炜, 张军平 . 步态识别的深度学习:综述[J]. 模式识别与人工智能, 2018,179(5):56-66.
	He Yiwei, Zhang Junping . Deep learning for gait recognition:A survey[J]. Pattern Recognition and Artificial Intelligence, 2018,179(5):56-66.
[2]	卢官明, 衣美佳 . 步态识别关键技术研究[J]. 计算机技术与发展, 2015(7):106-112.
	Lu Guanming, Yi Meijia . Research on critical techniques in gait recognition[J]. Computer Technology and Development, 2015(7):106-112.
[3]	范哲意, 蒋姣, 曾亚军 , 等. 基于空-频域特征和线性判别分析的视频步态识别[J]. 光学技术, 2017,43(4):374-380.
	Fan Zheyi, Jiang Jiao, Zeng Yajun , et al. Video human gait recognition based on space-frequency domain features and linear discriminant analysis[J]. Optical Technique, 2017,43(4):374-380.
[4]	王科俊, 丁欣楠, 邢向磊 , 等. 多视角步态识别综述[J]. 自动化学报, 2019,45(5):15-26.
	Wang Kejun, Ding Xinnan, Xing Xianglei , et al. A survey of multi-view gait recognition[J]. ACTA Automatica Sinica, 2019,45(5):15-26.
[5]	陈华, 曹真, 胡春海 . 动静态特征融合的正面视角步态识别[J]. 光电工程, 2013(11):87-92.
	Chen Hua, Cao Zhen, Hu Chunhai . Front-view gaitrecognition based on the fusion of static and dynamic features[J]. Opto-Electronic Engineering, 2013(11):87-92.
[6]	邵虹, 王昳昀 . 基于集成Gabor特征的步态识别方法[J]. 电子测量与仪器学报, 2017,31(4):573-579.
	Shao Hong, Wang Yiyun . Gait recognition method based on integrated gabor feature[J]. Journal of Electronic Measurement and Instrumentation, 2017,31(4):573-579.
[7]	史东承, 贾令尧, 梁超 , 等. 基于图像空间金字塔SURF-BoW的步态识别[J]. 计算机工程, 2017,43(9):270-275.
	Shi Dongcheng, Jia Lingyao, Liang Chao , et al. Gait recognition based on SURF-BoW of image space pyramid[J]. Computer Engineering, 2017,43(9):270-275.
[8]	Chen C, Liang J, Zhao H , et al. Frame difference energy image for gait recognition with incomplete silhouettes[J]. Pattern Recognition Letters, 2009,30(11):977-984.
[9]	Ahmed M, Al-Jawad N, Sabir A T. Gait recognition based on kinect sensor [C]. Brussels:Real-time Image & Video Processing. International Society for Optics and Photonics, 2014.
[10]	郇战, 陈学杰, 吕士云 , 等. 基于多分类器融合的步态识别方法[J]. 计算机应用, 2019,39(3):96-102.
	Huan Zhan, Chen Xuejie, Lü Shiyun , et al. Gait recognition method based on multiple classifier fusion[J]. Journal of Computer Applications, 2019,39(3):96-102.
[11]	Sebastijan S, Matjaz J . Inertial sensor-based gait recognition: A review[J]. Sensors, 2015,15(9):22089-22127.
[12]	蒋伟, 王瑞锦, 余苏喆 , 等. 基于步态识别的移动设备身份认证模型[J]. 电子科技大学学报, 2019,48(2):114-119.
	Jiang Wei, Wang Ruijin, Yu Suzhe , et al. Eesearch on identity authentication model of mobile devices based on gait recognition[J]. Journal of University of Electronic Science and Technology of China, 2019,48(2):114-119.
[13]	涂斌斌, 谷丽华, 揣荣岩 , 等. 基于MEMS加速度传感器的步态识别[J]. 中国惯性技术学报, 2017,25(3):304-308,359.
	Tu Binbin, Gu Lihua, Chuai Rongyan , et al. Gait recognition based on MEMS acceleration sensor[J]. Journal of Chinese Inertial Technology, 2017,25(3):304-308,359.
[14]	Choi S, Youn I H, Le May R . Biometric gait recognition based on wireless acceleration sensor using k-nearest neighbor classification[C]. Honolulu:IEEE International Conference on Computing, Networking and Communications ( ICNC), 2014.
[15]	Omid D, Mojtaba T, Raghvendar C V . IMU-based gait recognition using convolutional neural networks and multi-sensor fusion[J]. Sensors, 2017,17(12):2735-2745.
[16]	Smith J R, Fishkin K P, Jiang Bing , et al. RFID-based techniques for human-activity detection[J]. Communications of the ACM, 2005,48(9):39-46.
[17]	Tapia E M, Intille S S, Larson K. Activity recognition in the home using simple and ubiquitous sensors [C]. Berlin:Springer International Conference on Pervasive Computing, 2004.

	下衣口袋	胸前
正确率	88.12%	75.00%
错误接收率	1.48%	4.37%
错误拒绝率	1.48%	4.37%

训练样本数	10	20	30
正确率	80.50%	88.12%	91.87%
错误接收率	2.44%	1.48%	1.02%
错误拒绝率	2.44%	1.48%	1.02%