首页 | 本学科首页   官方微博 | 高级检索  
     

一种新的红外目标提取算法
引用本文:谢丽宽,李钊,胡双演,史德琴.一种新的红外目标提取算法[J].无线电工程,2009,39(6):26-27,33.
作者姓名:谢丽宽  李钊  胡双演  史德琴
作者单位:1. 海军驻西安地区导弹发动机军代室,陕西,西安,710025
2. 驻石家庄地区军事代表室,河北,石家庄,050081
3. 第二炮兵工程学院,陕西,西安,710025
摘    要:针对红外热图像普遍存在的目标与背景对比度差、边缘模糊等特点,提出了一种结合图像统计特征和模糊聚类的红外图像目标提取方法对红外图像进行迭代分块,通过计算统计特征度量函数的极值来确定种子区域,进而获得红外目标约束区域。利用模糊聚类对该约束区域进行有效的聚类划分,最终实现红外图像目标提取。实验结果表明,该方法能够准确的对红外目标进行提取,便于下一步的红外目标识别与跟踪。

关 键 词:红外图像  统计特征  模糊聚类  目标提取

A Novel Infrared Object Image Extraction Algorithm
XIE Li-kuan,LI Zhao,HU Shuang-yan,SHI De-qin.A Novel Infrared Object Image Extraction Algorithm[J].Radio Engineering of China,2009,39(6):26-27,33.
Authors:XIE Li-kuan  LI Zhao  HU Shuang-yan  SHI De-qin
Affiliation:XIE Li-kuan, LI Zhao, HU Shuang-yan, SHI De-qin(1. The Military Representative Office of NDMMP, Xi ' an Shanxi 710025, China; 2. The Military Representative Office in Shijiazhuang, Shijiazhuang Hebei 050081, China ; 3. The Second Artillery Engineering College, Xi ' an Shanxi 710025, China )
Abstract:A novel infrared object image extraction algorithm is proposed based on statistical feature and fuzzy clustering. Seed region of infrared object and contraint containing the entire object are obtained by performing iterative blocking of image and computing the extremes of statistical feature measure function. And then, object extraction is achieved by fuzzy clustering. Experimental results verify the effectiveness and robustness of this extraction method which can facilitate the target recognition and tracking in the next step.
Keywords:infrared image  statistical feature  fuzzy clustering  object extraction
本文献已被 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