结合自适应空间权重的改进型时空正则项跟踪算法

王鹏 1; 2; 孙梦宇 1; 王海燕 3; 李晓艳 1; 吕志刚 1

doi:10.7652/xjtuxb202005021

您当前的位置：

首页 >

文章列表页 >

结合自适应空间权重的改进型时空正则项跟踪算法

更新时间：2025-07-09

- 结合自适应空间权重的改进型时空正则项跟踪算法
- An Improved Spatial-Temporal Regulari-ation Term Tracking Algorithm Combining Adaptive Spatial Weights
- 西安交通大学学报 2020年54卷第5期页码：158-169
- 作者机构：
  
  1. 西安工业大学电子信息工程学院,西安,710021
  2. 西北工业大学航海学院,西安,710072
- 作者简介：
- 基金信息：
  
  : 2019-08-12。作者简介: 王鹏(1978—),男,教授。基金项目: 国家自然科学基金资助项目(61271362);陕西省科技厅重点研发计划资助项目(2019GY-022,2019GY-066)
- DOI：10.7652/xjtuxb202005021
  中图分类号： TP391
- 网络首发：2020-05-10，
  
  纸质出版：2020
- 稿件说明：
移动端阅览
王鹏 1, 2, 孙梦宇 1, 等. 结合自适应空间权重的改进型时空正则项跟踪算法[J]. 西安交通大学学报, 2020,54(5):158-169.

An Improved Spatial-Temporal Regulari-ation Term Tracking Algorithm Combining Adaptive Spatial Weights[J]. 2020, 54(5): 158-169.
王鹏 1, 2, 孙梦宇 1, 等. 结合自适应空间权重的改进型时空正则项跟踪算法[J]. 西安交通大学学报, 2020,54(5):158-169. DOI： 10.7652/xjtuxb202005021.

An Improved Spatial-Temporal Regulari-ation Term Tracking Algorithm Combining Adaptive Spatial Weights[J]. 2020, 54(5): 158-169. DOI： 10.7652/xjtuxb202005021.

摘要

为解决时空正则项的相关滤波视觉跟踪算法在目标部分遮挡时存在的模型漂移和尺度估计不准确问题

提出了结合自适应空间权重的改进型时空正则项跟踪算法。采用平均特征能量比将无法准确表达目标或过多表达背景信息的特征通道裁剪掉

以提高跟踪精度。在滤波器训练时加入空间权重正则项

利用时间正则项在目标遮挡时被动更新滤波器

使得在空间权重更新时更为准确

以此着重学习目标未被遮挡部分

获取可靠的相关滤波器系数。将滤波器求解划分为2个子问题

分别采用交替方向乘子法进行优化计算

保证算法运算速率。在牛顿迭代法中设置精度阈值

在保证定位精度的同时减少迭代次数。实验结果表明:在OTB-100数据集上所选择的6个视频序列中

所提算法较STRCF算法的平均中心位置误差降低了12.3像素

平均重叠率增加了7%

运算帧率可达19.25帧/s; 在OTB2015遮挡视频序列中

所提算法较STRCF算法的成功率曲线下积分面积(S

AUC

)增加了0.7%

使用深度特征的所提算法较DeepSTRCF和ASRCF算法的S

AUC

分别提升了3.9%与0.9%。

Abstract

To solve the problem of model drifting and inaccurate scale estimation for partially occluded object in spatial-temporal regulari-ed correlation filter visual tracking algorithm

an improved spatial-temporal regulari-ation term tracking algorithm combining adaptive spatial weights is proposed. For cutting out the feature channel that unables to express the background information more accurately

the average feature energy ratio channel is chosen to improve the tracking accuracy. Then the spatial weight regulari-ation term is added to the filter training

and the time regulari-ation term is used to passively update the filter when the target is occluded

so that the spatial weight update gets more accurate. Focusing on learning the non-occluded part of the target

the reliable correlation filter coefficient is obtained. The filter solution is divided into two subproblems

the alternating direction multiplier method is used for optimi-ation calculation to ensure the algorithm operation r

关键词

Keywords

references

王旭东, 王屹炜, 闫贺. 背景抑制直方图模型的连续自适应均值漂移跟踪算法 [J]. 电子与信息学报, 2019, 41(6): 1480-1487.

WANG Xudong, WANG Yiwei, YAN He. Continuously adaptive mean-shift tracking algorithm with suppressed background histogram mode [J]. Journal of Electronics & Information Technology, 2019, 41(6): 1480-1487.

张博, 江沸菠, 刘刚. 利用感知模型的长期目标跟踪 [J]. 中国图象图形学报, 2019, 24(11): 1906-1917.

ZHANG Bo, JIANG Feibo, LIU Gang. Long-term target tracking based on perceptual model [J]. Journal of Image and Graphics, 2019, 24(11): 1906-1917.

LI H M, SHI L P. Robust event-based object tracking combining correlation filter and CNN representation [J]. Frontiers in Neurorobotics, 2019, 13: 00082.

胡云层, 路红, 杨晨, 等. 融合DSST和KCF的尺度自适应跟踪算法 [J]. 计算机工程与设计, 2019, 40(12): 3563-3568.

HU Yunceng, LU Hong, YANG Chen, et al. Scale adaptive tracking algorithm combining DSST and KCF [J]. Computer Engineering and Design, 2019, 40(11): 3563-3568.

MEMARMOGHADAM A, MOALLEM P. Size-aware visual object tracking via dynamic fusion of correlation filter-based part regressors [J]. Signal Processing, 2019, 164: 84-98.

ISWANTO I A, CHOA T W, LI B. Object tracking based on meanshift and particle-Kalman filter algorithm with multi features [J]. Procedia Computer Science, 2019, 157: 521-529.

李牧子. 基于图像识别的目标检测与跟踪系统的设计与实现 [D]. 北京: 北京邮电大学, 2019: 40-41.

杨源, 库涛, 查宇飞, 等. 快速多特征金字塔的尺度目标跟踪方法 [J].西安交通大学学报,2016, 50(10): 49-56.

YANG Yuan, KU Tao, ZHA Yufei, et al. Fast multi-feature pyramid scale target tracking method [J]. Journal of Xi’an Jiaotong University, 2016, 50(10): 49-56.

LI Y, ZHU J K. A scale adaptive kernel correlation filter tracker with feature integration [C]∥Proceedings of the 13th European Conference on Computer Vision. Berlin, Germany: Springer, 2015: 254-265.

DANELLJAN M, HAGER G, KHAN F S, et al. Learning spatially regularized correlation filters for visual tracking [C]∥Proceedings of the 2015 IEEE International Conference on Computer Vision. Piscataway, NJ, USA: IEEE, 2015: 4310-4318.

DANELLJAN M, HAGER G, KHAN F S, et al. Convolutional features for correlation filter based visual tracking [C]∥Proceedings of the 2015 IEEE International Conference on Computer Vision Workshop. Piscataway, NJ, USA: IEEE, 2015: 7406433.

DANELLJAN M, BHAT G, KHAN F S, et al. ECO: efficient convolution operators for tracking [C]∥Proceedings of the 2017 IEEE Conference on Computer Vision and Pattern Recognition. Piscataway, NJ, USA: IEEE, 2017: 6931-6939.

LI F, TIAN C, ZUO W M, et al. Learning spatial-temporal regularized correlation filters for visual tracking [C]∥Proceedings of the 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Piscataway, NJ, USA: IEEE, 2018: 4904-4913.

DAI Kenan, WANG Dong, LU Huchuan, et al. Visual tracking via adaptive spatially-regularized correlation filters [C]∥Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. Piscataway, NJ, USA: IEEE, 2019: 4670-4679.

RAMDAS A, TIBSHIRANI R J. Fast and flexible ADMM algorithms for trend filtering [J]. Journal of Computational and Graphical Statistics, 2016, 25(3): 839-858.

DALAL N, TRIGGS B. Histograms of oriented gradients for human detection [C]∥Proceedings of the 2005 IEEE Computer Society Conference on Computer Vision and Pattern Recognition. Piscataway, NJ, USA: IEEE: 886-893.

DANELLJAN M, KHAN F S, FELSBERG M, et al. Adaptive color attributes for real-time visual tracking [C]∥Proceedings of the 2014 IEEE Conference on Computer Vision and Pattern Recognition. Piscataway, NJ, USA: IEEE, 2014: 1090-1097.

CHE M Q, WANG R L, LU Y, et al. Channel pruning for visual tracking [C]∥Proceeding of the 15th European Conference on Computer Vision. Berlin, Germany: Springer, 2019: 70-82.

WU Y, LIM J, YANG M H. Object tracking benchmark [J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2015, 37(9): 1834-1848.

WU Y, LIM J, YANG M H. Online object tracking: a benchmark [C]∥Proceedings of the 2013 IEEE Conference on Computer Vision and Pattern Recognition. Piscataway, NJ, USA: IEEE, 2013: 2411-2418.

吕明明,侯润民,柯于峰,等.光电跟踪平台脱靶量滞后补偿方法.2019,53(11):141-147.[doi:10.7652/xjtuxb201911 020]

廖加文,齐春,曹剑中,等.结合自适应特征选择和蕨类分类器的相关滤波跟踪算法.2019,53(6):101-108.[doi:10.7652/xjtuxb201906014]

侯利明,连峰,王伟.δ-广义标记多伯努利滤波器的非线性应用扩展.2019,53(6):109-116.[doi:10.7652/xjtuxb201906 015]

王树亮,毕大平,张奎,等.认知雷达波形和检测门限自适应跟踪算法.2019,53(4):65-71.[doi:10.7652/xjtuxb201904 010]

钟德星,杨元,刘瑞玲,等.基于单目视觉的装配机器人研究及应用.2018,52(5):81-87.[doi:10.7652/xjtuxb201805012]

彭华甫,黄高明,田威,等.标签多伯努利机动目标跟踪与分类算法.2019,53(2):157-162.[doi:10.7652/xjtuxb201902 021]

宋喜玉,任修坤,郑娜娥,等.多目标跟踪下的分布式MIMO雷达资源联合优化算法.2018,52(10):110-115.[doi:10.7652/xjtuxb201810015]

杜明洋,毕大平,王树亮,等.一种混合的机动群目标分离检测跟踪算法.2018,52(10):116-123.[doi:10.7652/xjtuxb 201810016]

隗寒冰,陈尧,贾志杰,等.融合历史轨迹的智能汽车城市复杂环境多目标检测与跟踪算法.2018,52(10):132-140.[doi:10.7652/xjtuxb201810018]

初红霞,谢忠玉,王科俊.一种结合颜色纹理直方图的改进型Camshift目标跟踪算法.2018,52(3):145-152.[doi:10.7652/xjtuxb201803020]

苗雨,宋骊平,姬红兵.箱粒子广义标签多伯努利滤波的目标跟踪算法.2017,51(10):107-112.[doi:10.7652/xjtuxb2017 10018]

庞策,黄树彩,刘锦昌,等.多传感器交叉提示技术在传感器联盟中的应用.2017,51(7):148-155.[doi:10.7652/xjtuxb 201707021]

浏览量

下载量

CSCD

文章被引用时，请邮件提醒。

提交

工具集

关联资源

一种多线索融合的均值偏移跟踪算法

结合自适应特征选择和蕨类分类器的相关滤波跟踪算法

离子液体改性燃料的电流体动力学效应研究

水蒸气掺混对氨裂解气旋流燃烧与排放特性的影响研究

采用试验设计方法的第三代CO2气体箔轴承箔片几何参数敏感性分析