Indoor person tracking method in considering interference noise perturbation

doi:10.15961/j.jsuese.2016.06.024

CORC > 兰州理工大学 > 兰州理工大学

	Indoor person tracking method in considering interference noise perturbation
	Li, Wei 2,3; Cao, Jie 1,2,3; Li, Jun 2,3; Wang, Jinhua 2,3
刊名	Sichuan Daxue Xuebao (Gongcheng Kexue Ban)/Journal of Sichuan University (Engineering Science Edition)
	2016-11-20
卷号	48 期号:6 页码:172-179
关键词	Color Covariance matrix Jacobian matrices Approximate filtering Color histogram H-infinity Linear approximations Observation equation Pedestrian tracking Real time performance Unscented transformations
ISSN号	10093087
DOI	10.15961/j.jsuese.2016.06.024
英文摘要	In order to solve indoor pedestrian tracking problem in the condition of light mutation and similar color interference, a new root mean square (RMS) unscented H-infinity filtering tracking method was proposed by considering noise perturbation interference. The unscented transform was used to replace Jacobian matrix computation, and the Gauss filter density was used to approximate filtering distribution, which reduced the linear approximation error generated in the observation equation. The RMS of target state covariance matrix was carried out by Cauchy decomposition, which weakened the system observation noise disturbing effect on observations, and reduced the computational complexity. The simulation results of similar color interference and light mutation showed that compared with the conventional UKF, EKF and PF nonlinear methods, the average tracking error was reduced by 30%, and the running time was reduced by 18.7%. The proposed method can effectively improve the accuracy of pedestrian tracking and enhance the real-time performance in the case of light mutation and similar color interference. © 2016, Editorial Department of Journal of Sichuan University (Engineering Science Edition). All right reserved.
语种	中文
出版者	Editorial Department of Journal of Sichuan University
内容类型	期刊论文
源URL	[http://ir.lut.edu.cn/handle/2XXMBERH/112974]
专题	兰州理工大学
作者单位	1.College of Computer and Communication, Lanzhou Univ. of Technol., Lanzhou; 730050, China 2.Gansu Manufacturing Info. Eng. Research Center, Lanzhou; 730050, China; 3.College of Electrical and Info. Eng., Lanzhou Univ. of Technol., Lanzhou; 730050, China;
推荐引用方式 GB/T 7714	Li, Wei,Cao, Jie,Li, Jun,et al. Indoor person tracking method in considering interference noise perturbation[J]. Sichuan Daxue Xuebao (Gongcheng Kexue Ban)/Journal of Sichuan University (Engineering Science Edition),2016,48(6):172-179.
APA	Li, Wei,Cao, Jie,Li, Jun,&Wang, Jinhua.(2016).Indoor person tracking method in considering interference noise perturbation.Sichuan Daxue Xuebao (Gongcheng Kexue Ban)/Journal of Sichuan University (Engineering Science Edition),48(6),172-179.
MLA	Li, Wei,et al."Indoor person tracking method in considering interference noise perturbation".Sichuan Daxue Xuebao (Gongcheng Kexue Ban)/Journal of Sichuan University (Engineering Science Edition) 48.6(2016):172-179.