| 摘要: |
| 对唐山湾三岛2010 年10 月10 m 分辨率的SPOT5 多光谱数据源, 采用不同分类方法进行识别和判断, 并对比不同分类器在遥感影像分类中的应用效果和分类精度。有针对性地探究海岛土地利用遥感分类过程中不同方法的优劣, 获取最适于岛陆地区土地利用的遥感分类方法。根据土地利用现状分类标准(GB/T21010-2007)和海岛陆域土地利用类型划分的相关规定, 将唐山湾三岛主要分为滩涂、裸地、林地、草地、居住区、内陆水体和潮间带(潮水覆盖区)6 类。并分别构建结合人为控制的非监督分类、监督分类和基于专家知识的决策树分类系统, 参照更高分辨率影像、先验知识和野外调查资料, 评价分类结果与实地调查结果的吻合程度, 最终通过总体分类精度和Kappa 系数等指标对各分类器精度进行评价和对比分析。 |
| 关键词: 唐山湾三岛 土地利用分类 遥感 监督分类 决策树分类 |
| DOI：10.11759/hykx20120401001 |
| 分类号: |
| 基金项目:海洋沉积与环境地质国家海洋局重点实验室开放基金资助项目(MASEG200807); 海洋公益性行业科研专项经费项目(200905004) |
|
| Comparison of land use classification methods used in analysis of the Tangshan Bay three islands using remote sensing images based on ENVI |
|
|
| Abstract: |
| In this paper, different classification methods were used to recognize the SPOT5 multispectral images of the Tangshan Bay three islands in October, 2010 with a resolution of 10 meters. The application and classification accuracy among different classifiers were compared. To find out the best remote sensing classification method for land usage analysis, we compared the advantages and disadvantages of all the methods used in land using classification analysis. According to the standards of land using classification and relevant rules about main land using classification of islands, we divided the Tangshan Bay three islands into tidal flat, bare land, forest, grass land, residential district, inland water and intertidal zone. Referring to high resolution images, future knowledge and field investigation data, we shave uccessfully set up the artificially controlled unsupervised classification and supervised classification and decision tree classification system based on expert knowledge. Then we evaluated the match condition between classification results and the results of actual investigation. At last, we evaluated and analyzed the accuracy of all the classifiers by some indicators such as Kappa. |
| Key words: the Tangshan Bay three islands land use classification remote sensing supervised classification decision tree classification |
