多特征融合匹配的多目标跟踪

闫辉; 许廷发; 吴青青; 徐磊; 吴威

doi:10.3788/CO.20130602.0163

多特征融合匹配的多目标跟踪

doi: 10.3788/CO.20130602.0163

北京理工大学光电学院光电成像技术与系统教育部重点实验室,北京100081

基金项目:

国家自然科学基金重点资助项目(科学仪器专项)(No.61027002);国家973重点基础研究发展计划资助项目(No.2009CB72400603);国家自然科学基金资助项目(No.60972100)

详细信息

作者简介:
闫辉(1986-),男,山东菏泽人,硕士研究生,2008年于北京航空航天大学获得学士学位,主要从事图像处理、目标跟踪等方面的研究。E-mail: yanhui17@sina.com;许廷发(1968—),男,黑龙江肇东人,博士,教授,博士生导师,1992年、2000年于东北师范大学获得学士、硕士学位,2004年于中国科学院长春光学精密机械与物理研究所获得博士学位,2006于华南理工大学电子与信息学院完成博士后工作,主要从事光电成像探测与识别等方面的研究。E-mail:ciom_xtf1@bit.edu.cn

通讯作者:
许廷发, E-mail:ciom_xtf1@bit.edu.cn

中图分类号: TP391.41
计量
- 文章访问数: 3463
- HTML全文浏览量: 630
- PDF下载量: 921
- 被引次数: 0
出版历程
- 收稿日期: 2012-11-17
- 修回日期: 2013-01-13
- 刊出日期: 2013-04-10

Multi-object tracking based on multi-feature joint matching

Key Laboratory of Photoelectronic Imaging Technology and System of the Ministry of Education, School of Optoelectronics, Beijing Institute of Technology, Beijing 100081, China

摘要

摘要: 针对复杂背景的视频图像多目标跟踪遮挡问题,提出了一种多特征融合匹配的多目标跟踪方法。基于自适应高斯混合背景模型重构和更新背景策略,实现当前帧背景减除和多目标检测;采用目标的颜色特征、质心位置、运动速度等特征进行融合匹配跟踪;最后,通过区域辅助判定策略将场景下的目标状态分为目标进入场景、目标退出场景、目标暂消、目标重现、目标融合和分裂5种状态,用质心预测方法和遮挡因子辅助匹配来提高匹配正确率。仿真实验结果表明:采用该方法跟踪同一目标和不同目标的相似度平均值分别为0.949 71和0.505 73,优于单一颜色特征信息匹配;目标遮挡结束后重新匹配相似度为0.972 83,实现了复杂背景下具有表面相似性的多目标实时跟踪。
- 多目标跟踪 /
- 高斯模型 /
- 多特征融合 /
- 质心预测 /
- 遮挡因子
Abstract: In order to solve the occlusion problem in multi-object tracking for the complex background of a video image, an approach for the multi-object tracking based on multi-feature joint matching is presented. First, the adaptive Gaussian mixture background model is used for reconstructing and updating the background to achieve the background subtraction of current frame and multi-object detection. Then, the joint matching tracking is developed based on matching color characteristics, positions and objects velocities. Finally, the objects in the scene are divided into entering, exiting, temporarily disappear of the object, the re-emergence of the object and the merge and split of the object, and the predicted position and the occlusion factor of the object are used to improve the accuracy of multi-feature joint matching. Experimental results indicate that the similarities of the same object and the different objects are 0.949 71 and 0.505 73 respectively in the tracking with the proposed approach, which is better than that of matching with the color characteristics. Furthermore, the similarity of object is 0.972 83 after occlusion. The approach is satisfactory for real-time tracking of multi-object with appearance similarity in a complex environment.
- multi-sobject tracking /
- Gaussian background model /
- multi-feature joint matching /
- predicted position /
- occlusion factor

HTML全文

参考文献(1)

[1] 顾鑫,王海涛,汪凌峰,等.基于不确定性度量的多特征融合跟踪[J].自动化学报,2011,37(5):550-559. GU X,WANG H T,WANG L F,et al.. Fusing multiple features for object tracking based on uncertainty measurement[J]. Acta Automatica Sinica,2011,37(5):550-559.(in Chinese) [2] 李安平.复杂环境下的视频目标跟踪算法研究[D].上海:上海交通大学,2006. LI A P. Research oil tracking algorithm for visual target under complex environments[D]. Shanghai:Shanghai Jiaotong University,2006.(in Chinese) [3] 郝志成,高文.多模跟踪技术在轮式侦察车图像处理器的应用[J].中国光学,2011,4(5):480-488. HAO ZH CH,GAO W. Application of multi-pattarn tracking technique in image processor of wheel type scout car[J]. Chinese Optics,2011,4(5):480-488.(in Chinese) [4] HARINI V,PAUL S,NIKOS P. Robust target detection and tracking through integration of motion, color, and geometry[J]. Computer Vision and Image Understanding,2006,103(2006):121-138. [5] 颜佳,吴敏渊.遮挡环境下采用在线Boosting的目标跟踪[J].光学精密工程,2012,20(2):439-446. YAN J,WU M Y. On-line boosting based target tracking under occlusion[J]. Opt. Precision Eng.,2012,20(2):439-446.(in Chinese) [6] 胡江龙,方景龙,王大全.多目标跟踪算法在水质监测中的应用[J].机电工程,2012,39(5):613-615. HU J L,FANG J L,WANG D Q. Water quality monitoring using multi-object tracking algorithm[J]. J. Mechanical Electrical Eng.,2012,39(5):613-615.(in Chinese) [7] 王坤峰,李镇江,汤淑明.基于多特征融合的视频交通数据采集方法[J].自动化学报,2011,37(3):322-330. WANG K F,LI ZH J,TANG SH M. Visual traffic data collection approach based on multi-features fusion[J]. Acta Automatica Sinica,2011,37(3):322-330.(in Chinese) [8] 丘文涛,赵建,刘杰.结合区域分割的SIFT图像匹配方法[J].液晶与显示,2012,20(2):439-446. QIU W T,ZHAO J,LIU J. Image matching algorithm combining SIFT with region segmentation[J]. Chinese J. Liquid Crystals and Displays,2012,20(2):439-446.(in Chinese) [9] STAUFFER C,GRIMSON W. Learning patterns of activity using real-time tracking[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence,2000,22(8):747-757. [10] NUMMIAFO K,KOLLER-MEIER E,GOOL L V. Color features for tracking non-rigid objects[J]. Acta Automatics Sinica,2003,29(3):345-355. [11] VIGNON D,LOVELL B C,ANDREWS R J. General purpose real-time object tracking using Hausdorff transforms[C]//9th International Conference on Information Processing and Management of Uncertainty in Knowledge-Based Systems,Annency,France,July 1-6,2002:1-6. [12] GARQI P. Robust multiple target tracking under occlusion using fragmented mean shift and Kalman filter[C]//2011 Internation Conference on Communications and Signal Processing(ICCSP),Kerala,India,Feb 10-12,2011:5171521. [13] PAPADAKIS N,BYGEAU A. Tracking with occlusions via graph cuts[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence,2011,33(1):144- 157.

施引文献

附加信息(0)

访问统计