[1]吴 刚,曾晓勤,苏守宝,等.融合增量主成分分析与粒子滤波的车辆表观模型跟踪[J].南京师范大学学报(自然科学版),2017,40(01):33.[doi:10.3969/j.issn.1001-4616.2017.01.006]
　Wu Gang,Zeng Xiaoqin,Su Shoubao,et al.Vehicle Appearance Model Tracker Integrating Incremental Principal Component Analysis and Particle Filter[J].Journal of Nanjing Normal University(Natural Science Edition),2017,40(01):33.[doi:10.3969/j.issn.1001-4616.2017.01.006]
点击复制

融合增量主成分分析与粒子滤波的车辆表观模型跟踪()
分享到：

《南京师范大学学报》（自然科学版）[ISSN:1001-4616/CN:32-1239/N]

卷:
第40卷
期数:
2017年01期
页码:
33
栏目:
·数学与计算机科学·
出版日期:
2017-03-31

文章信息/Info

Title:
Vehicle Appearance Model Tracker Integrating Incremental Principal Component Analysis and Particle Filter
文章编号:
1001-4616(2017)01-0033-06
作者:
吴 刚1; 2; 3; 曾晓勤3; 苏守宝1; 王池社1; 2
(1.金陵科技学院计算机工程学院,江苏 南京 211169)(2.金陵科技学院大数据研究院,江苏 南京 211169)(3.河海大学计算机与信息学院,江苏 南京 210098)
Author(s):
Wu Gang1; 2; 3; Zeng Xiaoqin3; Su Shoubao1; Wang Chishe1; 2
(1.School of Computer Engineering,JinLing Institute of Technology,Nanjing 211169,China)(2.Institute of Big Data,JinLing Institute of Technology,Nanjing 211169,China)(3.College of Computer and Information,Hohai University,Nanjing 210098,China)
关键词:
车辆跟踪; 表观模型; 自相关矩阵; 增量学习; 粒子滤波
Keywords:
vehicle tracking; appearance model; autocorrelation matrix; incremental learning; particle filter
分类号:
TP391.4
DOI:
10.3969/j.issn.1001-4616.2017.01.006
文献标志码:
A
摘要:
针对车辆运动方向持续变化、目标车辆距离远近变化、光照强度变化等场景下,稳定且实时性地跟踪车辆的难点问题,融合自相关矩阵增量主成分分析(Incremental Principal Component Analysis,IPCA)增量学习与粒子滤波算法的基础上,提出一种新的基于表观模型(Appearance Model,AM)的车辆跟踪方法,从跟踪初始利用自相关矩阵与特征值分解构建车辆的子空间图像,通过IPCA增量学习后的子空间均值、特征向量基共同参与似然概率密度的计算,提高粒子滤波算法粒子权值计算的精度. 标准视频的跟踪实验表明:对比P.Hall-IPCA与D.Ross-IPCA表观模型跟踪方法,所提AM-IPCA车辆跟踪方法将跟踪成功率分别由82.7%～92.3%、92.1%～95.2%提升至95.1%～96.4%.
Abstract:
Aiming at the difficulties on stably and timely tracking vehicle on the scenes such as volatile moving direction,varying pose and distance,illumination change,etc.,integrating autocorrelation matrix,incremental learning on IPCA and particle filter algorithm,new kind of vehicle tracking methods using appearance model is proposed. When beginning at original tracking time,the proposed method can timely learn the characteristic subspace images of vehicle,using autocorrelation matrix and eigen value decomposition. Based on IPCA incremental learning,likelihood probability density is computed on subspace mean and eigenvector,increasing computational precision on weights of particles on particle filter algorithm. The tracking results demonstrate that success tracking rate of the proposed AM-IPCA vehicle tracking method is raised to 95.1%～96.4%,compared with 82.7%～92.3% of P.Hall-IPCA and 92.1%～95.2% of D.Ross-IPCA appearance model tracking method.

参考文献/References:

[1] WALIA G S,KAPOOR R. Recent advances on multicue object tracking:a survey[J]. Artificial intelligence review,2016,46(1):1-39.
[2]杨欣,刘加,周鹏宇,等. 基于多特征融合的粒子滤波自适应目标跟踪算法[J]. 吉林大学学报(工学版),2015,45(2):533-539.
[3]HSIA C H,LIOU Y J,CHIANG J S. Directional prediction CamShift algorithm based on adaptive search pattern for moving object tracking[J]. Journal of real-time image processing,2016,12(1):183-195.
[4]SUN J,ZHANG S,ZHANG L. Object tracking with spatial context model[J]. IEEE transactions on signal processing letters,2016,23(5):727-731.
[5]MUTHUSWAMY K,RAJAN D. Particle filter framework for salient object detection in videos[J]. IET computer vision,2015,9(3):428-438.
[6]CEHOVIN L,LEONARDIS A,KRISTAN M. Visual object tracking performance measures revisited[J]. IEEE transactions on image processing,2016,25(3):1 261-1 274.
[7]CHOUVARDAS S,KOPSINIS Y,THEODORIDIS S. Robust subspace tracking with missing entries:the set-theoretic approach[J]. IEEE transactions on signal processing,2015,63(19):5 060-5 070.
[8]MIHAYLOVA L,CARMI A Y,SEPTIER F,et al. Overview of Bayesian sequential Monte Carlo methods for group and extended object tracking[J]. Digital signal processing,2013,25(1):1-16.
[9]YIN Y,XU D,WANG X,et al. Online state-based structured SVM combined with incremental PCA for robust visual tracking[J]. IEEE transactions on cybernetics,2015,45(9):1 988-2 000.

备注/Memo

备注/Memo:
收稿日期:2016-08-20.
基金项目:金陵科技学院高层次人才科研启动项目(jit-rcyj-201508)、国家自然科学基金项目(61375121)、国家自然科学基金项目(61305011)、南京市经信委项目(交通大数据公共服务平台)、南京市科委重大项目(大数据驱动下的大型客运枢纽监控预警与应急处置).
通讯联系人:吴刚,博士,副教授,研究方向:机器学习与智能系统. E-mail:zdhxwg@jit.edu.cn
更新日期/Last Update: 1900-01-01