Gesture Authentication Method Based on HMM and D-S Evidence Theory

doi:10.3969/j.issn.1671-1122.2018.10.003

Abstract

Abstract:

Aiming at the security problems of smart phone users’ privacy and information, a gesture recognition identity authentication method based on HMM and D-S evidence theory is adopted. Firstly, the original data of the gesture trajectory is collected by the touch screen sensor of the smart phone, the gesture feature sequence is preprocessed and extracted as the observation sequence of the HMM, and a legal user gesture model is established. Secondly, the preliminary judgment result is obtained by matching the test gesture feature sequence and the gesture model, and then use D-S evidence theory combination rules to get the fusion probability distribution function for the final certification decision. Simulation results show that compared with other methods, this method can effectively improve the accuracy of gesture recognition authentication.

Key words: gesture recognition, identity authentication, HMM, D-S evidence theory, mobile sensor

CLC Number:

TP309

Ziwen SUN, Fu LI. Gesture Authentication Method Based on HMM and D-S Evidence Theory[J]. Netinfo Security, 2018, 18(10): 17-23.

Figures/Tables 5

References 15

[1]	LAI Jianwei, ZHANG Dongsong.ExtendedThumb: A Target Acquisition Approach for One-handed Interaction with Touch-screen Mobile Phones[J]. IEEE Transactions on Human-Machine Systems, 2015, 45(3): 362-370.
[2]	NGUYEN T V, SAE-BAE N, MEMON N. DRAW-A-PIN: Authentication using Finger-drawn PIN on Touch Devices[EB/OL]. .
[3]	ZHAO Kaili, LI Danyi, LI Qiang, et al.Implementation of Identity Authentication Scheme Based on Smart Mobile Cryptographic Module[J]. Netinfo Security, 2017, 17(9): 107-110.
	赵凯利,李丹仪,李强,等. 基于智能移动终端密码模块的身份认证方案实现[J]. 信息网络安全,2017,17(9):107-110.
[4]	PRAMANA M I W, SUHARDI S, KURNIAWAN N B, et al. Keystroke Dynamics for Authentication using Dynamic Time Warping[C]//IEEE. 14th International Joint Conference on Computer Science and Software Engineering, July 12-14, 2017, Nakhon Si Thammarat, Thailand. New Jersey: IEEE, 2017: 1-5.
[5]	MA Guangkai, WU Ligang, WANG Yan. A General Subspace Ensemble Learning Framework via Totally-corrective Boosting and Tensor-based and Local Patch-based Extensions for Gait Recognition[EB/OL]. , 2018-1-22.
[6]	HU Qing, LIU Benyong.Speaker Recognition Algorithm Based on Convolutional Neural Networks[J]. Netinfo Security, 2016, 16(4): 55-60.
	胡青,刘本永. 基于卷积神经网络分类的说话人识别算法[J]. 信息网络安全,2016,16(4):55-60.
[7]	VIVARACHO-PASCUAL C, SIMON-HURTADO A, MANSO-MARTINEZ E, et al.Client Threshold Prediction in Biometric Signature Recognition by Means of Multiple Linear Regression and Its Use for Score Normalization[J]. Elsevier Science Inc, 2016, 55(C): 20-32.
[8]	MUHAMMAD P, DEVI S A. Hand Gesture User Interface for Smart Devices Based on Mems Sensors[EB/OL]. .
[9]	YANG C, HAN D K, KO H. Continuous Hand Gesture Recognition Based on Trajectory Shape Information[EB/OL]. .
[10]	GANESH S M, VIJAYAKUMAR P, DEBORAH L J.A Secure Gesture Based Authentication Scheme to Unlock the Smartphones[C]//IEEE. 2nd International Conference on Recent Trends and Challenges in Computational Models, February 3-4, 2017, Tindivanam, India. New Jersey: IEEE, 2017: 153-158.
[11]	BURIRO A, CRISPO B, DELFRARI F, et al.Hold and Sign: A Novel Behavioral Biometrics for Smartphone User Authentication[C]//IEEE. 2016 IEEE Security and Privacy Workshops, May 22-26, 2016, San Jose, CA, USA. New Jersey: IEEE, 2016: 276-285.
[12]	PANG Yongchun, SUN Ziwen, WANG Yao.Authentication Algorithm of Multi-touch Based on Mobile Touch Sensor[J]. Journal of Computer Applications, 2015, 35(6): 1780-1784.
	庞永春,孙子文,王尧. 基于手机触摸屏传感器多点触摸身份认证算法[J]. 计算机应用,2015,35(6):1780-1784.
[13]	MARTINEZ-DIAZ M, FIERREZ J, GALBALLY J.Graphical Password-based User Authentication with Free-form Doodles[J]. IEEE Transactions on Human-Machine Systems, 2016, 46(4): 607-614.
[14]	KUMAR P, GAUBA H, ROY P P, et al.Coupled HMM-based Multi-sensor Data Fusion for Sign Language Recognition[J]. Pattern Recognition Letters, 2017, 86(C): 1-8.
[15]	SUN Ziwen, LI Hui, JI Zhichegn.Fusion Image Steganalysis Based on Dempster-shafer Evidence Theory[J]. Control and Decision, 2011, 26(8): 1192-1196.
	孙子文,李慧,纪志成. 基于D-S证据理论的融合图像隐写分析[J]. 控制与决策,2011,26(8):1192-1196.

	M=1	M=2	M=3	M=4	M=5	M=6
N=1	7.59	5.36	3.46	4.68	3.78	2.68
N=2	6.56	4.78	3.69	3.87	3.82	2.47
N=3	5.34	4.02	2.33	2.88	3.24	1.99
N=4	3.88	2.58	2.96	1.96	2.71	2.06
N=5	2.12	3.26	2.18	2.56	2.11	1.60
N=6	2.32	2.17	2.11	2.02	1.85	1.79

	食指手势数据		中指手势数据
概率区间	m₁(ω_ture)/%	m₁(ω_false)/%	m₂(ω_ture)/%	m₂(ω_false)/%
[0.0, 4.0)	4.5	95.5	33.2	66.8
[4.0, 4.3)	15.6	84.4	1.5	98.5
[4.3, 4.6)	22.2	77.8	15.8	84.2
[4.6, 4.9)	87.5	12.5	85.6	14.4
[4.9, +∞)	97.4	2.6	98.7	1.3

	FRR/%		FAR/%
用户	食指	中指	食指	中指
1	4.25	4.25	2.87	3.13
2	3.95	3.39	1.85	2.63
3	3.53	4.53	2.23	3.04
4	2.86	3.76	1.79	1.98
5	2.94	4.27	2.14	3.26
6	3.67	3.23	2.38	2.55
均值	3.53	3.91	2.21	2.77

	FRR/%				FAR/%
用户	与融合	或融合	均值融合	本文方法	与融合	或融合	均值融合	本文方法
1	4.76	5.09	2.72	2.84	0.67	0.87	0.53	0.16
2	5.53	4.78	3.23	3.06	0.85	0.95	1.02	0.48
3	4.85	4.52	2.65	2.69	0.79	0.79	0.46	0.00
4	5.97	4.89	3.36	3.15	0.62	1.12	0.83	0.62
5	5.62	5.02	2.77	2.78	0.83	1.06	0.49	0.46
6	5.16	4.47	2.52	2.55	0.91	0.81	0.36	0.33
均值	5.32	4.79	2.38	2.85	0.78	0.95	0.62	0.34