基于Skewness纹理的北京城市化区域提取  

Extraction of Urbanized Area in Beijing Based on Skewness Texture

作　　者：李熙[1] 关泽群[1] 沈体雁[2] 

机构地区：[1]武汉大学遥感信息工程学院,湖北武汉430079 [2]北京大学政府管理学院,北京100871

出　　处：《安徽农业科学》2008年第17期7407-7409,7412,共4页Journal of Anhui Agricultural Sciences

摘　　要：北京地区部分农作物和城市化区域的光谱属性接近,仅仅利用光谱信息无法有效地提取北京的城市化区域。提出了利用Skewness纹理区分农作物和城市化区域,结合光谱和高程信息可以有效提取城市化区域。经过大量的样本检验表明,引入Skewness特征能够大幅度提高分类精度。In Beijing, spectrum of some cropland is very similar to that of urbanized area, so it is hard to extract the urbanized area only by spectrum. Skewness texture was proposed to separate cropland from urbanized area. Assisted by spectrum and elevation data, the urbanized area was extracted efficiently. Accuracy assessment based on many samples showed that Skewness could improve the classification accuracy significantly.

关 键 词：ETM+ 遥感 纹理 建成区 SKEWNESS 

分 类 号：S127[农业科学—农业基础科学]