| 基于粒子滤波的场景事件实时识别 |
| |
| 引用本文： | 肖小玲,李腊元.基于粒子滤波的场景事件实时识别[J].武汉理工大学学报(交通科学与工程版),2007,31(5):823-826. |
| |
| 作者姓名： | 肖小玲 李腊元 |
| |
| 作者单位： | 1. 武汉理工大学计算机科学与技术学院,武汉,430063;长江大学计算机学院,荆州,434023
2. 武汉理工大学计算机科学与技术学院,武汉,430063 |
| |
| 基金项目： | 国家自然科学基金;湖北省自然科学基金;湖北省教育厅科研项目 |
| |
| 摘 要： | 为解决智能监控场景中场景事件的检测与分析问题,提出了一种基于粒子滤波的场景事件实时识别方法.将场景事件分解为一系列的子事件,构成多层动态贝叶斯网络模型,模型中每一状态对应一种事件.采用Particle F ilter粒子滤波方法对模型中各节点状态的后验概率进行实时估计,以对场景事件进行实时识别.采样不同的粒子数目进行了对比仿真试验,仿真结果表明该方法能够得到较高的识别精度.
|
| 关 键 词： | 事件检测 场景事件 粒子滤波 动态贝叶斯模型 |
| 修稿时间： | 2007-06-04 |
Real Time Recognition of Situation Events Based on Particle Filter |
| |
| Xiao Xiaoling,Li Layuan.Real Time Recognition of Situation Events Based on Particle Filter[J].journal of wuhan university of technology(transportation science&engineering),2007,31(5):823-826. |
| |
| Authors: | Xiao Xiaoling Li Layuan |
| |
| Institution: | 1.School of Computer Science and Technology, WUT, Wuhan 430063;2.School of Computer Science, Yangtze University, Jinzhou 434023 |
| |
| Abstract: | To detect and analyze situation events in intelligent surveillance,a method of real time recognition of situation events based on particle filter is proposed.Situation event is decomposed into a sequence of sub events,and is modeled using a hierarchical dynamic Bayesian network,where each node denotes an event.The posterior probability of each node in a hierarchical dynamic Bayesian network is inferred based on particle filter in order to recognize situation events in real time.Simulating experiments with different sample numbers are made.The results show that better recognition precision can be achieved. |
| |
| Keywords: | event detection situation event particle filter dynamic Bayesian network |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
|
