基于栅格空间的地表覆盖精化分类方法研究

Research on Land Cover’s Refined Classification Method Based on Raster Space

提出了基于栅格空间分析的地表覆盖精化分类方法,分析了分类系统匹配、矢量数据栅格化、精化分类数据处理及精化分类结果,开发了精化分类软件,空间叠置分析算法采用与数据源栅格空间分辨率一致性无关的算法。以第一次全国地理国情普查成果中的林地、草地三级类为例,利用中国植被图(1∶100万)专题数据对其进行精化分类,分析了两者地表覆盖分类系统的定义与内涵;选取了试验区,基于栅格空间分析实现了精化分类;并采用野外实地调查的方式,对结果进行了精度验证。研究表明,利用栅格空间分析进行地表覆盖精化分类,能够形成反映多样化分类系统的地表覆盖精化分类数据产品,对第一次全国地理国情普查成果推广应用具有一定的实践意义。

Land cover’s refined classification method based on raster spatial analysis is proposed. It analyzes the method’s details of classification system matching, vector data rasterizing, data processing of refined classification, and result analyzing. During the data processing, it develops refined classification software whose spatial overlay analysis algorithm has nothing to do with the consistency of rasterizing spatial resolution of data source. It takes level Ⅲ classes of forest land and grassland obtained in the project of The First Geographic National Conditions Investigation for example, and refines it by utilizing the thematic data of Vegetation Map of China(1∶1 000 000). Mainly, it analyzes the definition and connotation of the two land cover classification systems. Significantly, it selects the test area, and realizes refined classification based on raster spatial analysis. Meanwhile, the accuracy verifying of the results is carried out by field surveying. The results show that the refined classification method based on raster spatial analysis can form diverse data products which integrate different land cover classification systems, and could display practical significance for the popularization and application of the achievements of The First Geographic National Conditions Investigation.