首页 | 本学科首页   官方微博 | 高级检索  
     

多特征实时电视跟踪器
引用本文:黄继昌,朱振福,阎兆春,邵苏苓,张凯,潘安君. 多特征实时电视跟踪器[J]. 系统工程与电子技术, 1992, 0(11)
作者姓名:黄继昌  朱振福  阎兆春  邵苏苓  张凯  潘安君
作者单位:航天航空工业部二院二○七所(黄继昌,朱振福,阎兆春,邵苏苓,张凯),航天航空工业部二院二○七所(潘安君)
摘    要:
本文描述一种多特征实时电视跟踪器,应用3×3中值滤波器对目标图像进行预处理,采用SOBEL算法提取目标的边缘信息,利用目标灰度和边缘联合特征的贝叶斯分类算法检测和分割目标,在低对比度小目标情况下可以稳定跟踪目标,经外场试验取得满意的结果,该电视跟踪器已用于地空武器系统中。

关 键 词:电视跟踪系统  目标跟踪  目标特征

Real-Time Multi-Features TV Tracker
Huang Jichang,Zhu Zhengfu,Yan Zhaochun Shao Shulin,Zhang Kai and Pan AnjunInstitute No. of the Second Academy,MAS. Real-Time Multi-Features TV Tracker[J]. System Engineering and Electronics, 1992, 0(11)
Authors:Huang Jichang  Zhu Zhengfu  Yan Zhaochun Shao Shulin  Zhang Kai  Pan AnjunInstitute No. of the Second Academy  MAS
Affiliation:Huang Jichang,Zhu Zhengfu,Yan Zhaochun Shao Shulin,Zhang Kai and Pan AnjunInstitute No.207 of the Second Academy,MAS
Abstract:
A real-time multi-features TV tracker is described. The target image is preprocessed by using a 3 x 3 median filter algorithm, the target edge information is extracted by using the SOBEL algorithm. According to the target grey-level and edge association features, the target can be detected and segmented by Bayesian classifier algorithm. The TV tracker can steadily track small targets of low contrast. Satisfactory results have been achieved in battlefield experiment.
Keywords:TV tracking   Multi-feature association   target detection.
本文献已被 CNKI 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号