| 卢柳叶,李光录,张莉,张伐伐.基于TM影像的半干旱区土地利用信息提取——以山西省定襄县为例[J].干旱地区农业研究,2012,30(1):217~223 |
| 基于TM影像的半干旱区土地利用信息提取——以山西省定襄县为例 |
| TM image-based land use information extraction in semi-arid region——A case study of Dingxiang County of Shanxi Province |
| |
| DOI：10.7606/j.issn.1000-7601.2012.01.37 |
| 中文关键词: SVM NDVI 纹理特征 土地利用 PCA |
| 英文关键词:SVM NDVI texture feature land use PCA |
| 基金项目:中国科学院知识创新工程重大项目（KSCX-YW-09-07）；中央高校基本科研业务费（QN2009040） |
|
| 摘要点击次数: 511 |
| 全文下载次数: 338 |
| 中文摘要: |
| 为了更准确提取半干旱区的土地利用信息，以山西省定襄县为研究区，利用Landsat TM影像，在主成分分析（PCA）的基础上，结合归一化植被指数（NDVI）和纹理信息，构建支持向量机（SVM）分类模型，提取研究区的土地利用信息。结果表明：与最大似然法和单纯依靠纹理特征SVM分类方法相比较，基于 NDVI和纹理特征的SVM分类法的分类精度有了显著提高，分类总精度达到了84.50%，Kappa系数为0.8113。研究表明，该方法对半干旱区的土地利用信息提取较为理想。 |
| 英文摘要: |
| In order to extract land use information in semiarid region efficiently, Dingxiang Country of Shanxi Province is taken as the research area. Based on Landsat TM image and PCA, a comprehensive model of SVM integrated NDVI with texture features is adopted to classify regional land use information by using model maker of ENVI software. The result shows: contrasted with the classification based on the method of Maximum Likelihood and Texture Features-based SVM, the accuracy of omprehensive classification model of SVM integrated NDVI with texture features is significantly improved, and the overall classification accuracy is 84.50%, while Kappa coefficient reaches 0.8113, this work has e
stablished a novel methodological framework for the extraction of land use information in semiarid region. |
| 查看全文 查看/发表评论 下载PDF阅读器 |