采用响应图置信区域自适应特征融合的相关滤波跟踪

高赟; 赵江珊; 罗久桓; 周浩

doi:10.3788/OPE.20192705.1178

您当前的位置：

首页 >

文章列表页 >

采用响应图置信区域自适应特征融合的相关滤波跟踪

信息科学 | 更新时间：2020-08-13

- 采用响应图置信区域自适应特征融合的相关滤波跟踪
- Adaptive feature fusion with the confidence region of a response map as a correlation filter tracker
- 光学精密工程 2019年27卷第5期页码：1178-1187
- 作者机构：
  
  1.云南大学信息学院，云南昆明 650504
  2.昆明物理研究所，云南昆明 650223
- 作者简介：
  
  [ "高赟 (1980-)，女，山西侯马人，博士，副教授，硕士生导师，2007年于西安电子科技大学获得硕士学位，2014年于云南大学获得博士学位，主要研究方向为视频运动目标跟踪。E-mail:gausegao@163.com" ]
  [ "赵江珊 (1995-)，女，云南普洱人，硕士研究生，2013年于楚雄师范学院获得学士学位，主要研究方向为视频运动目标跟踪。E-mail: missandai@163.com" ]
- 基金信息：
  
  国家自然科学基金青年项目资助(61802337);国家自然科学基金地区项目资助(11663007);中国博士后科学基金第62批面上项目资助(2017M623323XB);云南省应用基础研究面上项目资助(2016FB103)
- DOI：10.3788/OPE.20192705.1178
  中图分类号： TP391
- 收稿日期：2018-11-15，
  
  录用日期：2019-1-6，
  
  纸质出版日期：2019-05-15
- 稿件说明：
移动端阅览
高赟, 赵江珊, 罗久桓, 等. 采用响应图置信区域自适应特征融合的相关滤波跟踪[J]. 光学精密工程, 2019,27(5):1178-1187.

Yun GAO, Jiang-shan ZHAO, Jiu-huan LUO, et al. Adaptive feature fusion with the confidence region of a response map as a correlation filter tracker[J]. Optics and precision engineering, 2019, 27(5): 1178-1187.
高赟, 赵江珊, 罗久桓, 等. 采用响应图置信区域自适应特征融合的相关滤波跟踪[J]. 光学精密工程, 2019,27(5):1178-1187. DOI： 10.3788/OPE.20192705.1178.

Yun GAO, Jiang-shan ZHAO, Jiu-huan LUO, et al. Adaptive feature fusion with the confidence region of a response map as a correlation filter tracker[J]. Optics and precision engineering, 2019, 27(5): 1178-1187. DOI： 10.3788/OPE.20192705.1178.

摘要

针对固定特征融合权重的相关滤波跟踪算法在光照变化、目标形变下跟踪失败的问题

提出了采用响应图置信区域自适应特征融合的相关滤波跟踪算法以提高算法鲁棒性。首先，将响应图中高于响应图期望值的区域作为响应图置信区域，然后，根据HOG特征响应图置信区域计算出HOG特征响应图和颜色直方图特征响应图的融合权重，实现HOG和颜色直方图特征的自适应融合。仿真实验采用跟踪基准数据库(OTB-2015)中的100组视频序列进行实验，对比了5种流行的相关滤波跟踪算法。实验结果表明，本文算法的综合AUC和精度分别为0.609和0.814，尤其在光照环境下AUC和精度分别为0.655和0.847，相比Staple分别提升5.7%和5.6%。本文算法在光照和形变交叉环境下AUC达到0.681。在光照变化、目标形变、背景混乱、尺度变化等场景下，本文算法具备更优的跟踪性能。

Abstract

To correct the failure of some correlated filter trackers using a fixed weight feature fusion under illumination variation and deformation

the correlation filter tracker with an adaptive feature fusion and a confidence region of the response map is proposed for enhancing tracking robustness. The confidence region of the response map is the region where each response value is higher than the expectation of the response map. The fusing weights of a HOG response map and color histogram response map at every frame were calculated using the confidence region of the HOG response map

realizing adaptive fusion. The simulated experiments compared the proposed tracker with five popular correlation filter trackers using a benchmark video database

OTB-2015. The experimental results show that the AUC and precision were 0.609 and 0.814

respectively

whereas under OTB-2015 values of 0.655 and 0.847

respectively

were obtained. Under illumination variation

the obtained values were 5.7% and 5.6% higher than Staple

and the AUC was 0.681 under illumination variation and deformation. With illumination variation

target deformation

background clutter and scale variation

the proposed tracker exhibited better tracking performance than the previously developed methodologies.

关键词

Keywords

references

BOLME D S, BEVERIDGE J R, DRAPER B A, et al .. Visual object tracking using adaptive correlation filters[C]. IEEE conference Computer Vision and Pattern Recognition ( CVPR ), 2010 IEEE Conference on. IEEE , 2010: 2544-2550.

刘教民, 郭剑威, 师硕, 等.自适应模板更新和目标重定位的相关滤波器跟踪[J].光学精密工程, 2018, 26(8):2100-2111.

LIU J M, GUO J W, SHI SH, et al .. Correlation filter tracking based on adaptive learning rate and location refiner[J]. Opt. precision Eng. , 2018, 26(8):2100-2111. (in Chinese)

HENRIQUES J F, CASEIRO R, MARTINS P, et al .. High-speed tracking with kernelized correlation filters[J]. IEEE Transactions on Pattern Analysis & Machine Intelligence , 2015, 37(3):583-596.

葛宝义, 左宪章, 胡永江, 等.基于双步相关滤波的目标跟踪算法[J].红外与激光工程, 2018, 47(12):388-397.

GE B Y, ZUO X Z, HU YJ, et al .. Object tracking algorithm based two-step correlation filter[J] Infrared and laser engineering , 2018, 47(12): 388-397. (in Chinese)

DANELLJAN M, HAGER AND G, KHAN F S, et al .. Accurate scale estimation for robust visual tracking[C]. British Machine Vision Conference , 2014: 1-5.

杨德东, 毛宁, 杨福才, 等.利用最佳伙伴相似性的改进空间正则化判别相关滤波目标跟踪[J].光学精密工程, 2018, 26(2):492-502.

YANG D D, MAO N, YANG F C, et al .. Improved SRDCF object tracking via the Best-Buddies similarity[J]. Opt. precision Eng ., 2018, 26(2):492-502. (in Chinese)

张博, 江沸菠, 刘刚.利用视觉显著性和扰动模型的上下文感知跟踪[J].光学精密工程, 2018, 26(8):2112-2121.

ZHANG B, JIANG F B, LIU G, et al .. Context-aware tracking based on a visual saliency and perturbation model[J]. Opt. precision Eng ., 2018, 26(8):2112-2121. (in Chinese)

LI Y, ZHU J. A scale adaptive kernel correlation filter tracker with feature integration[C]. European Conference on Computer Vision, Springer, Cham , 2015: 254-265.

BERNARDINI C, SILVERSTON T, FESTOR O. MPC: Popularity-based caching strategy for content centric networks[C]. IEEE International Conference on Communications. , 2014: 3619-3623.

DANELLJAN M, KHAN F S, FELSBERG M, et al .. Adaptive color attributes for real-time visual tracking[C]. IEEE Conference on Computer Vision and Pattern Recognition ( CVPR ). IEEE , 2014: 1090-1097.

DANELLJAN M, HÄGER G, KHAN F S, et al .. Discriminative scale space tracking[J]. IEEE Transactions on Pattern Analysis & Machine Intelligence , 2017, 39(8):1561-1575.

HENRIQUES J F, RUI C, MARTINS P, et al .. Exploiting the circulant structure of tracking-by-detection with kernels[C]. Computer Vision - ECCV 2012. Springer Berlin Heidelberg , 2012: 702-715.

MA C, HUANG J B, YANG X, et al .. Hierarchical convolutional features for visual tracking[C]. IEEE International Conference on Computer Vision. IEEE Computer Society , 2015: 3074-3082.

DANELLJAN, MARTIN. Learning spatially regularized correlation filters for visual tracking[J]. IEEE International Conference on Computer Vision IEEE Computer Society , 2015:4310-4318.

WANG M, LIU Y, HUANG Z. Large margin object tracking with circulant feature maps[C]. IEEE conference Computer Vision and Pattern Recognition ( CVPR ), July 2017.

DANELLJAN M, BHAT G, KHAN F S, et al .. ECO: efficient convolution operators for tracking[J]. axXiv: 1611.09224v1, 2016:6931-6939.

BERTINETTO L, VALMADRE J, GOLODETZ S, et al .. Staple: complementary learners for real-time tracking[C]. IEEE Conference on Computer Vision and Pattern Recognition ( CVPR ). IEEE , 2016.

POSSEGGER H, MAUTHNER T, BISCHOF H. In defense of color-based model-free tracking[C]. IEEE conference Computer Vision and Pattern Recognition ( CVPR ), 2015: 2113-2120.

WU Y, LIM J, YANG M H. Object tracking benchmark[J]. IEEE Trans Pattern Anal Mach Intell , 2015, 37(9):1834-1848.

EVERINGHAM M, GOOL L V, WILLIAMS C K I, et al .. The pascal , visual object classes ( VOC ) Challenge [J]. International Journal of Computer Vision, 2010, 88(2):303-338.

浏览量

159

下载量

CSCD

文章被引用时，请邮件提醒。

提交

工具集

关联资源

基于知识蒸馏的Transformer视觉跟踪器

强化边缘特征提取的双分支融合真实图像去雾

针对RGBT跟踪的特殊属性的跨模态交互融合网络

基于混合时空卷积的轻量级视频超分辨率重建

联合自注意力和分支采样的无人机图像目标检测