|
|
论文:2013,Vol:31,Issue(5):746-752 |
|
|
引用本文： |
|
|
武君胜, 杨恒, 李阳. 基于粒子滤波框架下的自适应多特征融合目标建模算法[J]. 西北工业大学 |
|
|
Wu Junsheng, Yang Heng, Li Yang. Object Modeling Algorithm Based on Self-Adaptive Multi-Feature Fusion in Particle Filter[J]. Northwestern polytechnical university |
|
|
|
|
|
|
|
| 基于粒子滤波框架下的自适应多特征融合目标建模算法 |
|
| 武君胜1, 杨恒2, 李阳1 |
|
1. 西北工业大学 软件与微电子学院, 陕西 西安 710072;
2. 西安应用光学研究所, 陕西 西安 710065 |
| 摘要: |
| 针对单一特征描述目标模型的缺陷,提出了一种多特征融合粒子滤波跟踪算法,该算法采用具有互补性的灰度直方图特征和梯度直方图特征共同描述目标模型;在目标跟踪过程中,根据特征对目标和背景的区分程度,动态地调整每个特征的置信度,对目标模型进行在线动态建模和更新,以提高目标模型描述的准确度,并进一步提高粒子滤波算法的跟踪精度。新算法不仅可以满足目标跟踪的实时性要求,而且可避免特征的冗余,增加目标模型描述的多样性。实验结果验证了新算法的有效性。 |
| 关键词:
多特征融合
目标跟踪
粒子滤波
自适应
|
|
| Object Modeling Algorithm Based on Self-Adaptive Multi-Feature Fusion in Particle Filter |
|
| Wu Junsheng1, Yang Heng2, Li Yang1 |
|
1. Department of Software Engineering, Northwestern Polytechnical University, Xi'an 710072, China;
2. Xi'an Institute of Applied Optics, Xi'an 710065, China |
| Abstract: |
| A particle filter object tracking algorithm based on dynamic feature fusion is proposed in this paper. The presented algorithm uses the complementary features,which are gray histogram and gradient histogram,to represent the target model. In the tracking process,the confidence level for each feature is adjusted according to the degree of discrimination between the object and the background,and the target model is established and updated continuously online. The presented method can improve the accuracy of the target modeling and furthermore improve the accuracy of the particle filter tracking algorithm. The new algorithm not only meets the real-time requirements of the target tracking but also avoids the characterizing redundancy,thus increasing the diversity of the target model. Experimental results and their analysis demonstrate the effectiveness of our approach. |
| Key words:
algorithms
bandwidth
calculations
collision avoidance
computer vision
efficiency
error analysis
experiments
gradient methods
image fusion
mathematical models
redundancy
schematic diagrams
target tracking
dynamic feature fusion
object tracking
particle filter
self-adaptive
|
|
| 收稿日期: 2013-03-06
修回日期:
|
| DOI: |
| 基金项目: 陕西省自然科学基金(2011JM8006);航空科学基金(2009ZA53012)资助 |
| 通讯作者:
Email： |
| 作者简介: 武君胜(1962-),西北工业大学教授、博士生导师,主要从事软件工程和科学计算可视化等研究。
|
|
| 相关功能 |
|
|
|
|
| 作者相关文章 |
|
| 武君胜 在本刊中的所有文章 |
| 杨恒 在本刊中的所有文章 |
| 李阳 在本刊中的所有文章 |
|
|
|
|
|
|
|
|
参考文献: |
|
|
[1] Zhang B, Tian W, Jin Z.Probabilistic Tracking of Objects with Adaptive Cue Fusion Mechanism.Journal of Shanghai Jiaotong University, 2007, E-12(2): 185-190
[2] 高建坡, 王煜坚, 杨 浩, 吴镇扬.以颜色和形状直方图为线索的粒子滤波人脸跟踪.中国图象图形学报, 2007, 12 (3): 466-473 Gao Jianpo, Wang Yujian, Yang Hao, Wu Zhenyang.Particle Filter Face Tracking Using Color and Shape Histogram as Clues.Journal of Image and Graphics, 2007, 12(3): 466-473 (in Chinese)
[3] Fazli S, Pour H M, Bouzari H.Particle Filter Based Object Tracking with Sift and Color Feature.The Proceedings of the Second International Conference on Machine Vision, 2009, 89-93
[4] Li C, Wang X, Xu X.Robust Object Tracking with Adaptive Fusion of Color and Edge Strength Local Mean Features Based on Particle Filter.International Forum on Information Technology and Applications, 2009, 285-289
[5] 李远征, 卢朝阳, 高全学, 李 静.基于多特征融合的均值迁移粒子滤波跟踪算法.电子与信息学报, 2010, 32(2):411-415 Li Yuanzheng, Lu Zhaoyang, Gao Quanxue, Li Jing.Particle Filter and Mean Shift Tracking Method Based on Multi-Feature Fusion.Journal of Electronics & Information Technology, 2010, 32(2): 411-415 (in Chinese)
[6] 胡士强, 敬忠良.粒子滤波算法综述.控制与决策, 2005, 20(4): 361-365 Hu Shiqiang, Jing Zhongliang.Overview of Particle Filter Algorithm.Control and Decision, 2005, 20(4): 361-365 (in Chinese)
[7] 张旭光, 赵恩良, 王延杰.基于 Mean-Shift 的灰度目标跟踪新算法.光学技术, 2007, 33(2): 226-229 Zhang Xuguang, Zhao Enliang, Wang Yanjie.A New Algorithm for Tracking Gray Object Based on Mean-Shift.Optical Technique, 2007, 33(2): 226-229 (in Chinese)
[8] Peng Ningsong, Yamg Jie, Liu Zhi, et al.Automatic Selection of Kernel-Bandwidth for Mean-Shift Object Tracking.Journal of Software, 2005, 16(9): 1542-1550
[9] Nummiaro K, Koller-Meier E, Van L Gool.An Adaptive Color-Based Particle Filter.Image and Vision Computing, 2003, 21 (1): 99-110
[10] 马 丽, 常发亮, 乔谊正.基于均值漂移算法和粒子滤波算法的目标跟踪.模式识别与人工智能, 2006, 19 (6):787-793 Ma Li, Chang Faliang, Qiao Yizheng.Target Tracking Based on Mean Shift Algorithm and Particle Filtering Algorithm.Pattern Recognition and Artificial Intelligence, 2006, 19(6): 787-793 (in Chinese)
[11] 李红波, 曾德龙, 吴 渝.基于 Mean-Shift 和粒子滤波的两步多目标跟踪方法.重庆邮电大学学报, 2010, 22 (1):112-117 Li Hongbo, Zeng Delong, Wu Yu.A Two-Step Multiple Targets Tracking Algorithm Based on Mean-Shift and Particle Filter.Journal of Chongqing University of Posts and Telecommunications, 2010, 22(1): 112-117 (in Chinese)
[12] VSAM 项目测试视频数据库.http://www.vividevaluation.ri.cmu.edu/datasets/datasets.html VSAM Project Test Video Database.http://www.vividevaluation.ri.cmu.edu/datasets/datasets.html (in Chinese) |
|
|
|
相关文献: |
|
|
1．高社生, 薛丽, 魏文辉.渐消自适应Unscented粒子滤波及其在组合导航中的应用[J]. 西北工业大学, 2012,30(1): 27-31 |
|
|
|
|
|
|
