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

LANDSAT7ETM+图像的光谱模式分析与应用——以北京密云地区土地覆盖为例
引用本文:张香平,杨小唤.LANDSAT7ETM+图像的光谱模式分析与应用——以北京密云地区土地覆盖为例[J].地球信息科学,2005,7(1):107-110.
作者姓名:张香平  杨小唤
作者单位:中国科学院地理科学与资源研究所数据中心,北京 100101
摘    要:采用模式分解的方法(PDM),对北京密云地区2002年3月19日ETM+图像进行光谱模式提取与分解,得到研究区各个像元的植被、水和土壤系数;并通过不同系数的组合和判别分析的条件确定了研究区的土地覆盖状况以及不同土地覆盖类型的面积。结果表明,在降低ETM+数据维的情况下,光谱模式分解可以更直接的反映地物的类型。

关 键 词:模式分解法  标准光谱形状  分解系数  
收稿时间:2004-03-01;
修稿时间:2004年3月11日

Spectral Pattern Extract and Related Analysis Based on LANDSAT 7 ETM+ Images--Taking Land Cover in Miyun District of Beijing as a Case
ZHANG Xiangping,YANG Xiaohuan.Spectral Pattern Extract and Related Analysis Based on LANDSAT 7 ETM+ Images--Taking Land Cover in Miyun District of Beijing as a Case[J].Geo-information Science,2005,7(1):107-110.
Authors:ZHANG Xiangping  YANG Xiaohuan
Institution:Resources and Environmental Data Center, Institute of Geographic Sciences and Natural Resource Research,CAS, Beijing 100101, China
Abstract:The main purpose of this paper is to extract three spectral coefficients of water factor, vegetation factor and soil factor which are contained in every pixel of LANDSAT ETM+ images in order to get a new available method to process similar multi- or high-spectral images. It mainly applies a multi-spectral analysis method, i. e., Pattern Decomposition Method based on the LANDSAT ETM+ images at Beijing area. At first, this study transfers the DN value of the sensor recorded to the satellite reflectance and at the same time makes atmospheric correction to all six bands of 1-5 and 7. Secondly, through band math using Envi software to extract the standard spectral pattern: water pattern, vegetation pattern and soil pattern which are the common parts of the ground objects having the similar characters. Thirdly, decomposition is performed to the whole image using Mat-lab software to get the three decomposition coefficients for every pixel. Finally, a case applying the vicinity of Miyun Reservoir in Beijing, through the three coefficients, a direct land cover type and area are recognized and calculated based on some discrimination conditions. It is convenient to convert the resultant data to land cover grid, and based on the three coefficients ratio one can get the percentage of the objects in every pixel.
Keywords:pattern decomposition method(PDM)  standard spectral shape  decomposition coefficient
本文献已被 CNKI 维普 等数据库收录!
点击此处可从《地球信息科学》浏览原始摘要信息
点击此处可从《地球信息科学》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