融合模糊统计纹理特征的多线索粒子滤波跟踪

doi:10.13229/j.cnki.jdxbgxb20190789

摘要/Abstract

摘要：

针对粒子滤波跟踪算法使用单一特征鲁棒性差，以及粒子重采样策略易导致粒子退化、贫化等问题，提出了一种基于多特征、多线索的改进粒子滤波跟踪方法。使用引入邻域关系的Histon直方图描述目标的颜色特征，并建立了一种稳健的模糊统计纹理特征（FSTF）表达空间纹理信息，然后将其自适应地与区域颜色特征融合构建多线索的观测模型。在粒子滤波跟踪过程中，使用基于K-means的粒子权重聚类进行更为准确的后验分布估计。在重要性重采样阶段，保留高权重粒子的同时基于当前目标状态的先验分布产生新粒子，避免了粒子退化并保证了粒子的多样性。在标准测试集上的实验结果表明：相比其他基于粒子滤波框架的跟踪算法，本文方法能够得到更高的跟踪精度和成功率。与其他效果突出的流行跟踪算法相比，本文方法能在光照变化、目标形变和背景扰动场景下取得更好的跟踪效果。

关键词: 信息处理技术, 粒子滤波跟踪, Histon直方图, 模糊统计纹理特征, 多线索

Abstract:

In order to solve the problem that particle filter tracking algorithm uses single feature with lower robustness and the resample strategy is easy to cause particle degradation and impoverishment， an improved particle filter tracking method based on multi-feature and multi-cue is proposed. First， the Histon histogram that introduces the neighborhood relationship is used to describe the color characteristics of the target. Then a robust fuzzy statistical texture feature is used to express spatial texture information， and it is adaptively fused with regional color features to construct a multi-cue observation model. In the particle filter tracking process， K-means-based particle weight clustering is used for more accurate posterior distribution estimation. In the importance resample stage， high-weight particles are retained while new particles are generated based on the prior distribution of the current target state， which ensures particle diversity and avoids particle degradation. Experiments are carried out on the standard test set. Compared with other tracking algorithms based on particle filter framework， the proposed method obtains higher tracking accuracy and success rate. Compared with other popular tracking algorithms， the proposed method can achieve better tracking results under illumination variation， target deformation and background disturbance scenarios.

Key words: information processing technology, particle filter tracking, Histon histogram, fuzzy statistical texture feature, multi-cue

中图分类号:

TP911.72

金静,党建武,王阳萍,申东. 融合模糊统计纹理特征的多线索粒子滤波跟踪[J]. 吉林大学学报(工学版), 2021, 51(3): 1111-1120.

Jing JIN,Jian-wu DANG,Yang-ping WANG,Dong SHEN. Multi⁃cue particle filter tracking based on fuzzy statistical texture features[J]. Journal of Jilin University(Engineering and Technology Edition), 2021, 51(3): 1111-1120.

图/表 10

表1

纹理正确分类率比较 (%)"

GLCM			GLRLM			FTV
能量	相关性		长行程因子	行程总数百分比		模糊长行程因子 $R F 1$	模糊行程总数百分比 $R F 2$
82.3	86.6		87.5	89.5		89.8	92.3

表1

图1

图2

图3

表2

表3

图4

图5

图6

图7

参考文献 19

1	杨小军. 基于Unscented信息滤波器的分布式目标融合跟踪[J]. 吉林大学学报:工学版, 2015, 45(2): 658-662.
	Yang Xiao-jun. Distributed fusion and target tracking based on Unscented information filter[J]. Journal of Jilin University(Engineering and Technology Edition), 2015, 45(2): 658-662.
2	李娟, 刘晓龙, 卢长刚. 改进的粒子滤波重采样算法[J]. 吉林大学学报:工学版, 2015, 45(6): 2069-2074.
	Li Juan, Liu Xiao-long, Lu Chang-gang. Improvement of resampling algorithm of particle filter[J]. Journal of Jilin University(Engineering and Technology Edition), 2015, 45(6): 2069-2074.
3	Li S, Zhao S, Cheng B, et al. Lightweight particle filter for robust visual tracking[J]. IEEE Access, 2018, 6: 32310-32320.
4	Firouznia M, Faez K, Amindavar H, et al. Chaotic particle filter for visual object tracking[J]. Journal of Visual Communication and Image Representation, 2018, 53: 1-12.
5	Jenkins M D, Barrie P, Buggy T, et al. Selective sampling importance resampling particle filter tracking with multibag subspace restoration[J]. IEEE Transactions on Cybernetics, 2018, 48(1): 264-276.
6	Li J Q, Zhao R H, Chen J L, et al. Target tracking algorithm based on adaptive strong tracking particle filter[J]. Iet Science Measurement & Technology, 2016, 10(7): 704-710.
7	Zhai Y, Song P, Mou Z, et al. Occlusion-aware correlation particle filter target tracking based on RGBD data[J]. IEEE Access, 2018, 6: 50752-50764.
8	Lin Y, Huang D, Huang W. Target tracking algorithm based on an adaptive feature and particle filter[J]. Information, 2018, 9(6): 140-148.
9	Xin S, Yao H, Lu X. Dynamic multi-cue tracking using particle filter[J]. Signal Image & Video Processing, 2014, 8(1): 95-101.
10	Yin M, Jin Z, Sun H, et al. Multi-cue-based CamShift guided particle filter tracking[J]. Expert Systems with Applications, 2011, 38(5): 6313-6318.
11	Walia G S, Kapoor R. Robust object tracking based upon adaptive multi-cue integration for video surveillance[J]. Multimedia Tools & Applications, 2015, 75(23): 15821-15847.
12	Mohabey A, Ray A K. Rough set theory based segmentation of color images[C]∥Fuzzy Information Processing Society, Nafips International Conference of the North American, Atlanta, USA, 2000: 338-342.
13	Rami H, Hamri M, Masmoudi L. Objects tracking in images sequence using local binary pattern(LBP)[J]. International Journal of Computer Applications, 2013, 63(20): 19-23.
14	Graves D, Pedrycz W. Kernel-based fuzzy clustering and fuzzy clustering:a comparative experimental study[J]. Fuzzy Sets and Systems, 2010, 161(4): 522-543.
15	Wu Y, Lim J, Yang M H. Object tracking benchmark[J]. IEEE Transactions on Pattern Analysis & Machine Intelligence, 2015, 37(9): 1834-1848.
16	Zhang K, Zhang L, Yang M H. Real-Time compressive tracking[C]∥European Conference on Computer Vision, Florence, Italy, 2012: 864-877.
17	Kalal Z, Mikolajczyk K, Matas J. Tracking-learning-detection[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2012, 34(7): 1409-1422.
18	Ross D A, Lim J, Lin R S, et al. Incremental learning for robust visual tracking[J]. International Journal of Computer Vision, 2008, 77(1-3): 125-141.
19	Ji H, Ling H, Wu Y, et al. Real time robust L1 tracker using accelerated proximal gradient approach[C]∥IEEE Conference on Computer Vision & Pattern Recognition, Providence, RI, USA, 2012: 1830-1837.

相关文章 15

[1]	郭继昌,乔珊珊. 基于深度图的水下图像复原[J]. 吉林大学学报(工学版), 2021, 51(2): 677-684.
[2]	刘国华,周文斌. 基于卷积神经网络的脉搏波时频域特征混叠分类[J]. 吉林大学学报(工学版), 2020, 50(5): 1818-1825.
[3]	史再峰,李金卓,曹清洁,李慧龙,胡起星. 基于生成对抗网络的低剂量能谱层析成像去噪算法[J]. 吉林大学学报(工学版), 2020, 50(5): 1755-1764.
[4]	王柯俨,王迪,赵熹,陈静怡,李云松. 基于卷积神经网络的联合估计图像去雾算法[J]. 吉林大学学报(工学版), 2020, 50(5): 1771-1777.
[5]	谌华,郭伟,闫敬文,卓文浩,吴良斌. 基于深度学习的SAR图像道路识别新方法[J]. 吉林大学学报(工学版), 2020, 50(5): 1778-1787.
[6]	张薇,韩勇,金铭,乔晓林. 基于托普利兹矩阵集重构的相干信源波达方向估计[J]. 吉林大学学报(工学版), 2020, 50(2): 703-710.
[7]	程艳芬,姚丽娟,袁巧,陈先桥. 水下视频图像清晰化方法[J]. 吉林大学学报(工学版), 2020, 50(2): 668-677.
[8]	于晓辉,张志成,李新波,孙晓东. 基于状态空间模型的指数衰减正弦信号参数估计[J]. 吉林大学学报(工学版), 2019, 49(6): 2083-2088.
[9]	刘富, 权美静, 王柯, 刘云, 康冰, 韩志武, 侯涛. 仿蝎子振源定位机理的位置指纹室内定位方法[J]. 吉林大学学报(工学版), 2019, 49(6): 2076-2082.
[10]	马子骥,卢浩,董艳茹. 双通道单图像超分辨率卷积神经网络[J]. 吉林大学学报(工学版), 2019, 49(6): 2089-2097.
[11]	郭继昌,吴洁,郭春乐,朱明辉. 基于残差连接卷积神经网络的图像超分辨率重构[J]. 吉林大学学报(工学版), 2019, 49(5): 1726-1734.
[12]	曹运合,曾丽,王宇. 基于特征空间的子阵级自适应和差波束测角方法[J]. 吉林大学学报(工学版), 2019, 49(5): 1735-1744.
[13]	卢洋,王世刚,赵文婷,赵岩. 基于离散Shearlet类别可分性测度的人脸表情识别方法[J]. 吉林大学学报(工学版), 2019, 49(5): 1715-1725.
[14]	董超,刘晶红,徐芳,王仁浩. 光学遥感图像舰船目标快速检测方法[J]. 吉林大学学报(工学版), 2019, 49(4): 1369-1376.
[15]	王柯俨,胡妍,王怀,李云松. 结合天空分割和超像素级暗通道的图像去雾算法[J]. 吉林大学学报(工学版), 2019, 49(4): 1377-1384.

Metrics

Viewed

Full text

Abstract

Cited

Shared

Discussed

算法	IV	OCC	SV	DEF	MB	FM	IPR	OPR	OV	BC	LR	平均值
MSR	0.421	0.614	0.437	0.554	0.498	0.501	0.497	0.399	0.401	0.510	0.482	0.483
LPFT	0.454	0.432	0.659	0.568	0.565	0.492	0.508	0.453	0.385	0.442	0.351	0.482
Chaotic_PF	0.489	0.610	0.475	0.498	0.510	0.604	0.378	0.405	0.562	0.553	0.387	0.497
本文	0.568	0.431	0.465	0.671	0.542	0.548	0.540	0.451	0.398	0.652	0.451	0.519

算法	IV	OCC	SV	DEF	MB	FM	IPR	OPR	OV	BC	LR	平均值
MSR	0.511	0.693	0.604	0.644	0.588	0.614	0.541	0.426	0.540	0.600	0.581	0.577
LPFT	0.572	0.520	0.729	0.674	0.608	0.558	0.609	0.557	0.539	0.574	0.422	0.578
Chaotic_PF	0.504	0.677	0.531	0.510	0.611	0.697	0.415	0.528	0.651	0.559	0.457	0.558
本文	0.664	0.525	0.522	0.706	0.601	0.602	0.633	0.509	0.550	0.709	0.503	0.594