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

基于小波分解的星载SAR图像纹理信息提取
引用本文:胡召玲,郭达志,盛业华.基于小波分解的星载SAR图像纹理信息提取[J].遥感学报,2001,5(6):423-427.
作者姓名:胡召玲  郭达志  盛业华
作者单位:中国矿业大学测绘与空间信息工程研究所,
基金项目:国家自然科学基金资助项目(No.49871069)和博士点科研基金资助项目(No.97029005).
摘    要:论述利用小波变换提取合成孔径雷达(SAR)图像的多尺度纹理信息,借助Daubechies3正交小波,对图像进行小波分解,将小波变换各个频带输出的l1范数作为纹理特征值。选取徐州市区局部的Radarsat卫星SAR图像,提取出纹理图像,实验证明,该纹理提取方法能有效地提取出面目标的纹理信息。

关 键 词:SAR图像  小波变换  纹理  特征提取  合成孔径雷达
文章编号:1007-4619(2001)06-423-05
收稿时间:2000/10/18 0:00:00
修稿时间:2000年10月18

Extracting Textural Information of Satellite SAR Image Based on Wavelet Decomposition
HU Zhao-ling,GUO Da-zhi and SHENG Ye-hua.Extracting Textural Information of Satellite SAR Image Based on Wavelet Decomposition[J].Journal of Remote Sensing,2001,5(6):423-427.
Authors:HU Zhao-ling  GUO Da-zhi and SHENG Ye-hua
Abstract:Texture extraction from SAR image is an important frontier field because SAR image has characteristics of high spatial resolution and abundant texture information. Wavelet transformation can provide an effective tool for multi scale texture analysis of SAR image. This paper describes a method to extract multi scale textural information from satellite SAR image using wavelet transformation. Firstly the SAR image is decomposed using Daubechies3 orthogonal wavelet and l 1 norms of every band image are computed as feature value of image pixels. According to the feature value on every channel, texture images which respectively denote texture on different scale of SAR image are produced. The method has been applied for the extraction the textural informtion from a Radarsat image of Xuzhou city. The experiment shows that the facial target textural information can be effectively extracted using the mehtod.
Keywords:SAR  image  wavelet transformation  texture  feature extraction  
本文献已被 CNKI 维普 万方数据 等数据库收录!
点击此处可从《遥感学报》浏览原始摘要信息
点击此处可从《遥感学报》下载免费的PDF全文
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