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

一种基于粗糙度估计的红外小目标检测算法
引用本文:肖利平,陶文兵,田金文,黄中非.一种基于粗糙度估计的红外小目标检测算法[J].华中科技大学学报(自然科学版),2002,30(10):65-68.
作者姓名:肖利平  陶文兵  田金文  黄中非
作者单位:1. 华中科技大学图像识别与人工智能研究所,图像信息处理与智能控制教育部重点实验室
2. 武汉市排水设施,有偿使用管理处
基金项目:武器装备预研基金资助项目,北京大学视觉与听觉信息处理国家重点实验室开放课题基金资助项目(No.2 0 0 2 0 6)
摘    要:根据红外图像的特点,提出了一种基于红外图像背景粗糙度的自适应小目标检测快速算法,该算法首先通过对背景的粗糙度的估计来自适应调节LOG算子的两个关键参数;高斯分布因子σ及强度因子k的值,用经过参数调整后的LOG算子进行边缘检测,检测出目标的大致轮廓,根据目标轮廓定出目标的中心点(即种子点),然后由中心点开始进行区域生长,最后得到比较理想的目标分割图像,实验结果表明了该算法的有效性及实用性。

关 键 词:红外小目标检测算法  边缘检测  粗糙度估计  区域生长  LOG算子  高斯分布因子  强度因子  红外图像
文章编号:1671-4512(2002)10-0065-04
修稿时间:2002年5月29日

An algorithm of small targer detection based on the roughness estimation
Xiao Liping Tao Wenbing Tian Jinwen Huang Zhongfei,Postgraduate.An algorithm of small targer detection based on the roughness estimation[J].JOURNAL OF HUAZHONG UNIVERSITY OF SCIENCE AND TECHNOLOGY.NATURE SCIENCE,2002,30(10):65-68.
Authors:Xiao Liping Tao Wenbing Tian Jinwen Huang Zhongfei  Postgraduate
Institution:Xiao Liping Tao Wenbing Tian Jinwen Huang Zhongfei Postgraduate, Institute of Pattern Recognition & Artificial Intelligence,Huazhong Univ. of Sci. & Tech.,Wuhan 430074,China.
Abstract:According to the character of intrared images, this paper puts forward an adaptive small target detection algorithm, which is based on the estimation of the background roughness. The system can adaptively adjust the two parameters of LOG operator, the space constant of the Gaussian and the scale factor by estimating the background roughness. The two adjusted parameters are used to perform the edge detection and get the approximate contour of the object. The center of the object (the so-called seed) is calculated in terms of this calculated contour. As the region growth from the center of the object contour develops, some good segmented object images are produced. The results of the experiment show that this algorithm is efficient and practical.
Keywords:edge detection  roughness estimation  region growth  target detection
本文献已被 CNKI 维普 万方数据 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