首页 | 本学科首页   官方微博 | 高级检索  
     检索      

红外序列图像中运动小目标的检测方法
引用本文:李国宽,彭嘉雄,李红.红外序列图像中运动小目标的检测方法[J].华中科技大学学报(自然科学版),1999(11).
作者姓名:李国宽  彭嘉雄  李红
作者单位:华中理工大学模式识别与人工智能研究所图像信息处理与智能控制国家教育部重点实验室(李国宽,彭嘉雄),华中理工大学数学系(李红)
摘    要:研究红外图像中运动小目标的检测和跟踪问题.根据红外图像中背景干扰和噪声的特点,采用梯度倒数加权低通滤波器对图像进行滤波.在目标分割和标记后,沿目标轨迹进行能量积累以搜索和跟踪小目标.实验结果表明,该方法能极大地提高实际红外目标图像的信噪比,改善目标分割效果,能有效地进行小目标的检测和跟踪,尤其适用于目标遮掩或偶尔出现目标丢失的情况.

关 键 词:图像识别  小目标检测和跟踪  图像预处理  红外图像序列

Detection Method of Small Moving Targets in IR Sequence
Li Guokuan,Peng Jiaxiong,Li Hong.Detection Method of Small Moving Targets in IR Sequence[J].JOURNAL OF HUAZHONG UNIVERSITY OF SCIENCE AND TECHNOLOGY.NATURE SCIENCE,1999(11).
Authors:Li Guokuan  Peng Jiaxiong  Li Hong
Institution:Li Guokuan Peng Jiaxiong Li Hong
Abstract:A new algorithm is presented for detecting and tracking small moving targets in IR image sequences. Corresponding with characteristics of background clutter and noise, the gradient inverse weighted neighborhood averaging (GIWNA) filter is adopted. The energy accumulation along possible trajectories is performed to detect and track targets. Experiment results with real IR image sequence show that the method can greatly improve SNR and target segmentation, and detect and track small moving targets efficiently. It is especially suitable for the cases where a target is partly occluded or missing occasionally.
Keywords:image recognition  small target detection and tracking  image pre  processing  IR image sequence
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