基于目标灰度的土地利用变化信息边界提取  

Extracting Edges of Land Using Change Information with Objects Gray Value

作　　者：严丽娟[1] 冯德俊[2] 李永树[2] 

机构地区：[1]成都理工大学环境与土木工程学院,四川成都610059 [2]西南交通大学地理信息工程中心,四川成都610031

出　　处：《现代测绘》2005年第5期6-9,共4页Modern Surveying and Mapping

基　　金：国家自然科学基金资助项目(40371098)

摘　　要：为了提高大区域土地利用变化信息提取效率,本文在变化信息自动发现的基础上,基于目标灰度信息,研究遥感影像中土地利用变化信息的边界提取方法。由于研究对象是大区域中的小目标,无法直接利用灰度直方图寻找分割目标的灰度分布范围。本文提出了一种典型区域法,方法是利用区域中分割目标密集的多个典型小区域的灰度直方图计算所有目标的分割阈值。由于影像中地物众多,为了减少噪声的影响,提高计算速度,先利用双阈值分割法对目标进行粗分割,再对其进行小波变换,利用其模值,以及结合目标灰度信息,可以将目标边界快速地提取出来。In order to improve efficient of extracting land-using change information for large area, on the base of automatic detection of land-using change information, this paper study on the method of extraction edges for land-using change information in remote sensing image with its gray information. Since extracted targets are little objects in large area, gray value range of extracted targets couldn,t compute directly with gray histogram of large area image. We bring forward a typical area method, which uses gray histogram of several typical little areas whose targets are dense to compute segmentation threshold for all targets. The objects are numerous in remote sensing image. In order to reduce noise and improve computing speed, the objects are segmented roughly with segmentation method of dual-threshold, and computed results with wavelet, then object edges are extracted rapidly with wavelet transform module value and objects gray information.

关 键 词：遥感 边界提取 小波变换 土地利用 典型区域法 

分 类 号：P237[天文地球—摄影测量与遥感] P285.239[天文地球—测绘科学与技术]