基于OSM的地表覆盖变化数据提取

2014年,中国研制出了2000年和2010年两期30 m分辨率的全球地表覆盖数据产品,为了保持该数据集的现势性,持续更新成为该领域关键问题。变化数据的快速收集是全球地表覆盖数据更新的核心,志愿者地理信息(volunteered geographic information,VGI)的兴起为解决此问题提供了一条新的途径。在目前VGI项目中,OpenStreetMap(OSM)具有数据丰富、现势性强等优点,可以作为全球地表覆盖数据更新的参考信息来源,然而OSM数据不能直接对应地表覆盖增量数据,因此,本文提出了一种基于OSM的地表覆盖增量数据提取方法。该方法首先以OSM定义的几何类型和属性数据作为分类依据,建立OSM数据到地表覆盖数据的转换规则;然后针对类型转换后数据存在的目标零散等问题,采用四叉树聚类方法对转换数据进行综合;继而对整合后的数据栅格化并与GlobeLand30数据进行叠加运算得到地表覆盖增量数据。最后以爱沙尼亚区域的OSM数据验证了本文方法的可行性。

Land Cover Change-only Data Extraction Based on OSM Data

The global land cover data products were developed by China, with a resolution of 30m in 2000 and 2010. In order to maintain the currency of the dataset, continuous updating has become a key problem. The rapid collection of change-only data is one of the core issues of global land cover data updating. The development of volunteer geographic information (VGI) has provided a new way to solve this problem. In current VGI projects, OpenStreetMap (OSM) has the advantages of abundant data and strong currency, and can be used as a reference source for the updating of global land cover data. However, the OSM data does not directly correspond to the land cover incremental data. Based on the above analysis, the authors presented an incremental datadata extraction method for land cover data using OSM data source in this paper. Firstly, the OSM-data to land-cover-data conversion rules were established based on the geometric and attribute data defined by OSM features. Secondly, in order to remove the gap among the OSM transformed objects, a quadtree clustering method was used to aggregate the transformed data. Thirdly, the aggregated data was rasterized and compared with the existing GlobeLand30 datasets to obtain the land cover incremental data. Finally, relative data in Estonian was used to verify the adatability of this method.