摘要
本文首先讨论以图像块运动估计目标运动方向,然后分析Kalm an滤波器估测目标运动状态的原理、方法与过程。最后进行多目标视频监控实验。结果表明,本文研究的多目标跟踪技术实时性、准确性、稳定性好,灵敏度高,可用于视频监控业务。
This paper discusses estimating object-movable-orientation by graph-block-movement firstly, and then predicts principles, methods and processes of object movable status by analysing Kalman filter. Finally experiment of multi-object-video-supervising is carried out. The result proves that multi-object tracing technology researched by this paper has real-time performance, accuracy, stability and flexibility. So that it can be used in video supervisory business.
出处
《计算机应用与软件》
CSCD
北大核心
2005年第11期22-23,135,共3页
Computer Applications and Software
基金
江苏省高新技术产业发展项目基金资助(苏发改高技发[2004]672号)。
