地理国情普查中地表覆盖信息提取技术研究

地理国情普查的开展,促使地表覆盖信息提取技术向智能化、高效化发展;以西藏地区作为研究区域,选用2011年11月空间分辨率为0.5 m的WorldView_2影像作为数据源,采用目视解译、计算机自动解译、计算机人工交互解译3种遥感信息提取方法进行土地覆盖信息提取;通过实验对比分析,3种解译各具优势与不足,分类结果精度与遥感影像质量与解译区地形复杂程度和目标地物种类息息相关,在地理国情普查过程中需要综合使用.

Research on Land Cover Information Extraction Technology in Geographical Condition Survey

The geographical condition survey boosts land cover information extraction technology to realize intelligent and highly-efficient development. By taking Tibet area as research region, by taking WorldView_2 image with 0.5 m spatial resolution rate in November, 2011 as data source, by taking such three kinds of remote sensing information extraction methods as visual interpretation, computer automatic interpretation and computer-human interactive interpretation to extract land cover information, via experiment comparison and analysis, the research finds that there are different advantages and disadvantages among the three interpretation methods, and that there is a close correlation between classification result accuracy and remote image quality and complex extent of the landform in the interpreted areas and species information used in the process of geographical condition survey. in the targeted areas, which should be comprehensively