互联网 qkzz.net
全刊杂志网:首页 > 女性 > 文章正文
刊社推荐

一种用于智能监控的目标检测和跟踪方法


□ 王素玉 沈兰荪 李晓光

  摘 要:在对现有目标检测、跟踪算法进行分析对比的基础上,设计并实现了一种简单有效的目标检测和跟踪方案。首先提出了一种基于像素灰度归类和单模态高斯模型的背景重构算法,能够利用多帧包含前景目标的场景图像重构准确的背景模型。进而以此为基础采用背景减法进行各帧中目标的检测,并选取形心作为匹配特征实现了场景中多个目标的有效跟踪。实验表明,该方法实现简单,无须事先提供背景图像即可实现目标的准确检测和跟踪,其性能明显优于传统基于时间平均背景模型的方法。
  关键词:目标检测; 目标跟踪; 背景重构; 高斯模型; K均值聚类
  中图分类号:TP391 文献标志码:A 文章编号:1001-3695(2008)08-2393-03
  
  Object detection and tracking method for intelligent surveillance
  WANG Su-yu1,2, SHEN Lan-sun1, LI Xiao-guang1
  (1.Signal & Information Processing Lab, Beijing University of Technology, Beijing 100022, China; 2.College of Electronic & Information Engineering, Hebei University, Baoding Hebei 071002, China)
  Abstract:This paper designed and realized a simple and effective object detection and tracking scheme, based on a review of the existed detection and tracking algorithms. At first, proposed a pixel intensity classification and the single Gaussian model based background reconstruction algorithm, which could provide an accurate background model through a sequence of scene images with foreground objects. Then used the background subtraction method for object detection, selected the center of the object as the matching feature for tracking of multi-objects among the sequence. Experimental results show that the proposed algorithm and scheme is simple to realize, and can detect and track the moving objects effectively, it shows obvious perfor-mance improvement compared with the traditional time-averaged background based method. ......
很抱歉,暂无全文,若需要阅读全文或喜欢本刊物请联系《计算机应用研究》杂志社购买。
欢迎作者提供全文,请点击编辑
分享:
 

了解更多资讯,请关注“木兰百花园”
分享:
 
精彩图文


关键字
支持中国杂志产业发展,请购买、订阅纸质杂志,欢迎杂志社提供过刊、样刊及电子版。
关于我们 | 网站声明 | 刊社管理 | 网站地图 | 联系方式 | 中图分类法 | RSS 2.0订阅 | IP查询
全刊杂志赏析网 2017