[1]陈春雨,阳秋光,李东方.基于三维坐标的运动目标跟踪[J].应用科技,2018,45(02):23-28.[doi:10.11991/yykj.201612007]
 CHEN Chunyu,YANG Qiuguang,LI Dongfang.Moving objective tracking based on 3D coordinates[J].yykj,2018,45(02):23-28.[doi:10.11991/yykj.201612007]
点击复制

基于三维坐标的运动目标跟踪(/HTML)
分享到:

《应用科技》[ISSN:1009-671X/CN:23-1191/U]

卷:
第45卷
期数:
2018年02期
页码:
23-28
栏目:
现代电子技术
出版日期:
2018-03-15

文章信息/Info

Title:
Moving objective tracking based on 3D coordinates
作者:
陈春雨 阳秋光 李东方
哈尔滨工程大学 信息与通信工程学院, 黑龙江 哈尔滨 150001
Author(s):
CHEN Chunyu YANG Qiuguang LI Dongfang
College of Information and Communication Engineering, Harbin Engineering University, Harbin 150001, China
关键词:
背景差法sift算法三维重构FCM聚类Kalman滤波三维坐标二维坐标立体匹配
Keywords:
background subtraction methodsift algorithmthree-dimensional reconstructionFCM clusteringKalman filteringthree-dimensional coordinatestwo-dimensional coordinatesstereo matching
分类号:
TN957
DOI:
10.11991/yykj.201612007
文献标志码:
A
摘要:
基于二维坐标的多运动目标跟踪,在跟踪过程中由于目标相互遮挡,算法无法分清各个运动目标,导致跟踪目标失败。而三维坐标具有深度信息,利用目标遮挡前后坐标的不突变性能很好地分清各个目标,为此提出基于三维坐标的运动目标跟踪方法。首先,采用背景差法进行目标检测;其次使用sift算法对目标特征提取,运用极线约束对目标特征点进行立体匹配以及三维重构并使用模糊C均值聚类算法(FCM),确定运动目标中心三维坐标;最后结合Kalman滤波实现目标跟踪。实验和分析结果表明,算法能够较好地适应目标遮挡下的跟踪,具有良好的准确性、鲁棒性。
Abstract:
As for multi-objective tracking based on two-dimensional coordinates, because of mutual objective obstruction, the algorithm is hard to distinguish from moving objectives and the objective tracking may fail; while three-dimensional coordinates have depth information, the moving objectives can be clearly identified due to the non-mutation of coordinates before and after objective obstruction, therefore, the paper proposed a new multi-objective tracking algorithm based on three-dimensional coordinates. Firstly, the background subtraction method is used to detect objectives; secondly, sift algorithm is used to extract features of objectives, the epipolar constraint is utilized to carry out stereo matching and the three-dimensional reconstruction for the objective feature points, in addition, fuzzy C-means algorithm(FCM) clustering is used to determine the three-dimensional coordinates of the center of moving multi-objectives; finally, Kalman filtering method is combined with to implement objective tracking. Experimental and analytical results show that the proposed algorithm can properly adapt to objective tracking in case of obstruction, it has good robustness and accuracy.

参考文献/References:

[1] 谭艳, 王宇俊. 一种结合背景差分的改进CamShift目标跟踪方法[J]. 西南师范大学学报:自然科学版, 2016, 41(9):120-125.
[2] 翟明. 基于特征匹配与卡尔曼预测的多目标跟踪算法研究[D]. 沈阳:东北大学, 2014:1-55.
[3] KRIEGER E, SIDIKE P, ASPIRAS T, et al. Directional ringlet intensity feature transform for tracking[C]//Proceedings of 2015 IEEE International Conference on Image Processing. Quebec, Canada:IEEE, 2015:3871-3875.
[4] BARBIERI T TDS, GOULARTE R. KS-SIFT:A keyframe extraction method based on local features[C]//Proceedings of 2014 IEEE International Symposium on Multimedia. Taichung, China:IEEE, 2014:13-17.
[5] JI Wenping, GUO Baolong, DING Guiguang. Detecting traffic volume statistics based on virtual coil with optical flow method[J]. Computer simulation, 2004, 21(1):109-110, 117.
[6] ASARI V K, ARIGELA S. Color image enhancement of low-resolution images captured in extreme lighting conditions[C]//Proceedings of SPIE 9120, Mobile Multimedia/Image Processing, Security, and Applications. Baltimore:SPIE, 2014:91200Q.
[7] 姚放吾, 许辰铭. 基于目标质心的Meanshift跟踪算法[J]. 计算机技术与发展, 2012, 22(6):104-106, 110.
[8] 郭文强, 高晓光侯, 勇严. 基于卡尔曼预测算法的云台三维空间目标跟踪[J]. 陕西科技大学学报, 2007, 25(3):89-92.
[9] 余旺盛, 田孝华, 侯志强, 等. 基于关键区域特征匹配的视觉跟踪算法[J]. 电子学报, 2014, 42(11):2150-2156.
[10] SRIHARSHA K V, RAO N V. Dynamic scene analysis using Kalman filter and mean shift tracking algorithms[C]//Proceedings of 20156th International Conference on Computing, Communication and Networking Technologies. Denton:IEEE, 2016:1-8.
[11] 曲海琴. 一种改进型运动目标检测算法及其在视频监控系统中的应用研究[D]. 长春:吉林大学, 2013:1-47.
[12] GIBSON J D, KOO B, GRAY S D. Filtering of colored noise for speech enhancement and coding[J]. IEEE transactions on signal processing, 1991, 39(8):1732-1742.

相似文献/References:

[1]白鹤,郑丽颖.基于SIFT和三角网格的VHR建筑物检测技术[J].应用科技,2017,44(06):48.[doi:10.11991/yykj.201610001]
 BAI He,ZHENG Liying.Building detection from VHR based on SIFT and Delaunay triangulation[J].yykj,2017,44(02):48.[doi:10.11991/yykj.201610001]

备注/Memo

备注/Memo:
收稿日期:2016-12-07。
基金项目:国家自然科学基金项目(81913080)
作者简介:陈春雨(1974-),男,博士,副教授;阳秋光(1989-),男,硕士研究生
通讯作者:阳秋光,E-mail:526254174@qq.com
更新日期/Last Update: 2018-04-09