首页 | 本学科首页   官方微博 | 高级检索  
     

红外图像模式识别的预处理
引用本文:张燕妮,娄树理.红外图像模式识别的预处理[J].兵工自动化,2006,25(9):42-43,49.
作者姓名:张燕妮  娄树理
作者单位:1. 大连医科大学,计算机教研室,辽宁,大连,116027
2. 海军航空工程学院,控制工程系,山东,烟台,264001
摘    要:红外图像模式识别系统,由图像预处理、特征提取及模式分类组成.基于红外图像模式识别的图像预处理,包括滤波降噪、目标增强、图像分割等处理.滤波降噪有邻域平均、中值和Butterworth低通等滤波方法.目标增强可通过微分及Butterworth高通滤波完成,图像分割可采用阈值分割技术实现.故综合比较各种处理方法,采用两次均值滤波及拉普拉斯算子、阚值分割技术,较好实现了红外图像模式识别的预处理.

关 键 词:红外图像  模式识别  图像预处理  红外  图像模式识别  预处理  Pattern  Recognition  Infrared  Image  阚值  拉普拉斯算子  均值滤波  处理方法  比较  综合  分割技术  阈值  高通滤波  微分  滤波方法  低通  Butterworth  邻域平均  图像分割
文章编号:1006-1576(2006)09-0042-02
收稿时间:2006-02-10
修稿时间:2006-02-102006-06-13

Preprocessing for Infrared Image Pattern Recognition
ZHANG Yan-ni,LOU Shu-li.Preprocessing for Infrared Image Pattern Recognition[J].Ordnance Industry Automation,2006,25(9):42-43,49.
Authors:ZHANG Yan-ni  LOU Shu-li
Affiliation:1. Staff Room of Computer, Dalian Medical University, Dalian 116027, China; 2. Dep. of Control Engineering, Naval Aeronautical Engineering Institute, Yantai 264001, China
Abstract:The infrared pattern recognition system consists of image preprocessing, feature extraction and pattern classification. The image preprocessing based on infrared image mode recognition includes noise reduction, image enhancement and image segmentation. The filter noise reduction method includes neighborhood average method, median method and Butterworth low-pass method. The target enhancement could be realized through the differential coefficient and Butterworth high pass filter; the image division could be realized by adopting threshold division technology. So, compared with varied processing method, the two-mean filter, the laplacian operator and the threshold division technology could realize the infrared image pattern recognition oreprocessing better.
Keywords:Infrared image  Pattern recognition  Image preprocessing
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