基于加权重叠率的单目标视觉跟踪评价指标

doi:10.3969/j.issn.0372-2112.2017.03.036

PDF(1846 KB)

电子学报 ›› 2017, Vol. 45 ›› Issue (3) : 753-761. DOI: 10.3969/j.issn.0372-2112.2017.03.036

学术论文

基于加权重叠率的单目标视觉跟踪评价指标

孙巧, 张胜修, 张正新, 曹立佳, 李小锋

作者信息 +

A Weighted-Overlap Based Metric for Single Visual Object Tracking Evaluation

SUN Qiao, ZHANG Sheng-xiu, ZHANG Zheng-xin, CAO Li-jia, LI Xiao-feng

Author information +

文章历史 +

摘要

针对真值标注的歧义性、偏差性问题和具有缩放场景的视觉跟踪应用情况，提出了一种新的视觉跟踪单目标基准评价指标.首先，在重叠率基础上提出了加权重叠率框架；其次，提出了多区域标注方法，通过多区域标注降低标注者歧义性，在具有缩放场景的应用中，通过反演进行多区域标注，使评价更符合应用实际；再次，针对标注的偏差性，提出了多标注融合方法，提高了标注的可信度；最后，将应用于单次跟踪评价的重叠率框架推广到多次跟踪评价，利用加权结果图使评价更具解释性.通过著名评价标准VOT、OTB的真值标注融合实验验证了本文标注规则的准确性；通过在具有缩放场景的视觉跟踪实验和重复实验，与其他跟踪指标的比较验证了本文指标的有效性.

Abstract

Aimed at the problems of annotation of ground truth and the application of zooming,a new basic metric for visual tracking evaluation is proposed.Firstly,a weighted-overlap frame is reconstructed based on the traditional overlap.Secondly,we put forward multiple region annotation to decrease the deviation and apply in zooming.Thirdly,a multi-label fusion method is presented to improve the confidence level of the labels.Last but not least,the presented methods are expanded to repeated visual tracking evaluation,where a weighted result chart is utilized to make the evaluation more explanatory.Experimental results show that our annotation rule are more accurate than VOT and OTB,and the proposed metric is more appropriate than other metric.

导出引用

孙巧, 张胜修, 张正新, 曹立佳, 李小锋. 基于加权重叠率的单目标视觉跟踪评价指标[J]. 电子学报, 2017, 45(3): 753-761. https://doi.org/10.3969/j.issn.0372-2112.2017.03.036

SUN Qiao, ZHANG Sheng-xiu, ZHANG Zheng-xin, CAO Li-jia, LI Xiao-feng. A Weighted-Overlap Based Metric for Single Visual Object Tracking Evaluation[J]. Acta Electronica Sinica, 2017, 45(3): 753-761. https://doi.org/10.3969/j.issn.0372-2112.2017.03.036

中图分类号： TP391.4

参考文献

[1] 李康,何发智,潘一腾,等.基于簇相似的多分类器目标跟踪算法[J].电子学报,2016,44(4):821-825. LI Kang,HE Fa-zhi,PAN Yi-teng,et al.Multi-classifier object tracking based on cluster similarity[J].Acta Electronica Sinica,2016,44(4):821-825.(in Chinese)
[2] 宋涛,李鸥,刘广怡.基于置信区域内多级动态层表达的类贯序蒙特卡洛视觉跟踪方法[J].电子学报,2016,44(6):1355-1361. SONG Tao,LI Ou,LIU Gang-yi.Quasi-sequential Monte Carlo visual tracking based on multilevel dynamic layer representations in confidence region[J].Acta Electronica Sinica,2016,44(06):1355-1361.(in Chinese)
[3] PANG Y,LING H.Finding the best from the second bests-inhibiting subjective bias in evaluation of visual tracking algorithms[A].Proceedings of the 2013 IEEE International Conference on Computer Vision[C].Sydney,NSW,Australia:IEEE Press,2013.2784-2791.
[4] SMEULDE AWM,CHU DM,CUCCHIARA R,et al.Visual tracking:an experimental survey[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2014,36(7):1442-1468.
[5] WU Y,LIM J,YANG M.Object tracking benchmark[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2015,37(9):1834-1848.
[6] NAWAZ T,CAVALLARO A.A protocol for evaluating video trackers under real-world conditions[J].IEEE Transactions on Image Processing,2013,22(4):1354-1361.
[7] KRISTAN M,PFLUGFELDER R,LEONARDIS A,et al.The visual object tracking VOT2013 challenge results[A].Proceedings of the 2013 IEEE International Conference on Computer Vision Workshops[C].Sydney,NSW,Australia:IEEE Press,2013.98-111.
[8] KRISTAN M,MATAS J,LEONARDIS A,et al.A novel performance evaluation methodology for single-target trackers[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2016,38(11):2137-2155.
[9] TENG H,YU L,LIU H,et al.A novel performance evaluation method for visual tracking methods[A].Proceedings of the 2015 Chinese Intelligent Automation Conference[C].Fuzhou China:Springer,2015.313-324.
[10] CEHOVIN L,LEONARDIS A,KRISTAN M.Visual object tracking performance measures revisited[J].IEEE Transactions on Image Processing,2015,25(3):1261-1274.
[11] KASTURI R,GOLDGOF D,SOUNDARARAJAN P,et al.Framework for performance evaluation of face,text,and vehicle detection and tracking in video:Data,metrics,and protocol[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2009,31(2):319-336.
[12] 王国华,刘琼,庄家俊.基于局部特征的车载红外行人检测方法研究[J].电子学报,2015,43(7):1444-1448. WANG Guo-hua,LIU Qiong,ZHU Jia-jun.Method research on vehicular infrared pedestrian detection based on local features[J].Acta Electronica Sinica,2015,43(7):1444-1448.(in Chinese)
[13] DANELLJAN M,SHAHBAZ KHAN F,FELSBERG M,et al.Adaptive color attributes for real-time visual tracking[A].IEEE Conference on Computer Vision and Pattern Recognition[C].USA:IEEE Press,2014.1090-1097.
[14] JIA X,LU H,YANG M.Visual tracking via adaptive structural local sparse appearance model[A].Proceedings of the 2012 IEEE Conference on Computer Vision and Pattern Recognition[C].Providence,RI,USA:IEEE Press,2012.1822-1829.
[15] HENRIQUES JF,CASEIRO R,MARTINS P,et al.High-speed tracking with kernelized correlation filters[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2015,37(3):583-596.
[16] SEVILLA-LARA L,LEARNED-MILLER E.Distribution fields for tracking[A].Proceedings of the 2012 IEEE Conference on Computer Vision and Pattern Recognition[C].Providence,RI,USA:IEEE Press,2012.1910-1917.
[17] WANG D,LU H,YANG M.Online object tracking with sparse prototypes[J].IEEE Transactions on Image Processing,2013,22(1):314-325.
[18] ZHANG K,ZHANG L,LIU Q,et al.Fast visual tracking via dense spatio-temporal context learning[A].Proceedings of the 2014 European Conference on Computer Vision[C].Florence,Italy,2014.127-141.
[19] 余旺盛,田孝华,侯志强,等.基于局部分块学习的在线视觉跟踪[J].电子学报,2015,43(1):74-78. YU Wang-sheng,TIAN Xiao-hua,HOU Zhi-qiang,et al.Online visual tracking based on local patch learning[J].Acta Electronica Sinica,2015,43(1):74-78.(in Chinese)