J4 ›› 2009, Vol. 36 ›› Issue (3): 443-447.

• 研究论文 • 上一篇    下一篇

多传感器协同探测目标的分类识别研究

胡国平1,2;张守宏2   

  1. (1. 西安电子科技大学 雷达信号处理重点实验室,陕西 西安  710071;
    2. 空军工程大学 导弹学院,陕西 三原  713800)
  • 收稿日期:2008-11-03 修回日期:2008-12-15 出版日期:2009-06-20 发布日期:2009-07-04
  • 通讯作者: 胡国平
  • 基金资助:

    “863”计划资助(2006AA701304)

Study of target recognition based on multi-sensor cooperating with exploration

HU Guo-ping1,2;ZHANG Shou-hong1
  

  1. (1. Key Lab. of Radar Signal Processing, Xidian Univ., Xi'an  710071, China;
    2. Missile Inst. of Air Force Eng. Univ., Xi'an  713800, China)
  • Received:2008-11-03 Revised:2008-12-15 Online:2009-06-20 Published:2009-07-04
  • Contact: HU Guo-ping

摘要:

采用可见光、红外、SAR、高光谱、电子信号5类传感器对目标进行协同探测,提出了一种基于神经网络与D-S证据理论结合、采用“分层有序”进行目标分类识别的方法,该方法先对目标进行粗分类,然后进行细分类,能准确有效地进行目标的分类识别.

关键词: 协同探测, 分层有序, 神经网络, D-S理论, 目标识别

Abstract:

Applying a Multi-sensor, such as optical, infrared, SAR, High-spectrum and electron signal type, to explore the target cooperating, a new approach is proposed based on the Neural Network together with the D-S evidence theory, and “orderly quantization” is adopted in target recognition. The target is classified extensively and then intensively. The approach can accurately and effectively identify the target.

Key words: cooperating with exploration, orderly quantization, neural network, D-S evidence theory, target recognition

中图分类号: 

  • TP751.1