联合人体模型与块生长的人群目标分割

邓颖娜; 朱虹; 刘薇; 张晓丹

doi:10.3724/SP.J.1146.2009.00119

联合人体模型与块生长的人群目标分割

doi: 10.3724/SP.J.1146.2009.00119

计量
- 文章访问数: 3256
- HTML全文浏览量: 111
- PDF下载量: 745
- 被引次数: 0
出版历程
- 收稿日期: 2009-01-21
- 修回日期: 2009-07-21
- 刊出日期: 2010-03-19

Human Pose Model and Block Growth Combined Crowd Segmentatio

摘要

摘要: 人群目标的准确分割，是进行多相机联合目标跟踪与识别的关键。该文首先构造包含位置、尺寸、姿态信息的人体粗略姿态模型，并利用贝叶斯模型获得对应的目标模型。然后将前景区域利用分水岭算法划分为颜色分布一致的子块，通过位置和颜色约束解决遮挡目标的种子块选取问题。最后通过块生长获得每个目标区域。对于颜色相似子块，通过比较其产生的边缘能量确定其所属目标。实验结果表明，本文能够实现对人群目标较精确分割，且对背景噪声具有一定的抗干扰能力。
- 贝叶斯模型; 姿态模型; 目标分割; 块生长; 分水岭算法
Abstract: Crowd object segmentation is a key issue of the object tracking and recognition in multiple cameras. Human rough models with position, scale and pose information are constructed and then get the corresponding models by using Bayesian model. Then, the foreground is segmented into blocks of similar color distribution. Then the problem of the seed blocks selection is solved thought of color and position information under human inter-occlusion, and human region is received by seed growth. For blocks with similar color, they are merged into the objects by comparing the edge energy they brought. It can be seen that the method could segment the crowd precisely, and is not sensitive to background noise from the experimental results.

HTML全文

参考文献(1)

Yang Ran and Zheng Qin-fen. Multi-moving people detectionfrom binocular sequences[C]. Proceedings of InternationalConference on Acoustic, Speech and Signal Processing. HongKong, 2003, Vol.2: 297-300.[2]Brostow G J and Cipolla R. Unsupervised Bayesian detectionof independent motion in crowds[C]. Proceedings of IEEEComputer Society Conference on Computer Vision andPattern Recognition, New Yourk: IEEE Computer Society,2006, Vol.1: 594-601.[3]Elgammal A M and Davis LS. Probabilistic framework forsegmenting people under Occlusion[C]. Proceedings of 8thIEEE Conference on Computer Vision,Vancouver, 2001,Vol 2:145-152.[4]Ramanan D and Forsyth D A. Finding and tracking peoplefrom the bottom up[C]. Proceedings of IEEE Conference onComputer Vision and Pattern Recognition, Madison, 2003,Vol.2: 467-474.[5]Zhe Lin, Davis L S, and Doermann D, et al.. An interactiveapproach to pose-assisted and appearance-basedsegmentation of humans[C]. Proceedings of IEEE 11thInternational Conference on Computer Vision, Rio de Janeiro,Brazil, 2007: 1-8.[6]Wu Bo and Nevatia R. Detection and tracking of multiple,partially occluded humans by Bayesian combination ofedgelet based part detectors[J]. International Journal ofComputer Vision, 2007, 27(2): 247-266.[7]Zhao Tao and Ram Nevatia. Bayesian human segmentationin crowded situations[C]. Proceedings of IEEE ComputerSociety Conference on Computer Vision and PatternRecognition, Madison, 2003, Vol.2: 459-466.[8]Zhe Lin, Davis L S, Doermann D, and Dementhon D.Hierarchical part-template matching for human detectionand segmentation[C]. Proceedings of IEEE 11th InternationalConference on Computer Vision, Rio de Janeiro, 2007: 1-8.[9]杨莉, 杨新. 基于区域划分的曲线演化多目标分割[J]. 计算机学报, 2004, 27(3): 420-425.Yang Li and Yang Xin. Multi-object segmentation based oncurve evolving and region division[J]. Chinese Journal ofComputers, 2004, 27(3): 420-425.[10]Jacky S C, Yuk K K, and Wong R, et al.. Real-time multiplehead shape detection and tracking system with decentralizedtrackers[C][J].Proceedings of 6th IEEE InternationalConference on Intelligent System Design and Applications,Jinan.2006, Vol.2:384-389

施引文献

资源附件(0)

访问统计