摘要
以ALOS数据作为卫星遥感数据信息源,利用计算机进行监督分类和人机交互式解译,可提取工程区实际退耕还林面积,但对所有图斑进行实地植被状况调查的费用是非常昂贵和费时的。采用不等概抽样方法(PPS),将给定的遥感解译图斑面积/各工程实施图斑面积作为辅助因子,根据30块工程实际面积和GPS实地量测退耕还林面积,估计工程保存造林总面积,将两者提取面积做了对比分析。分析结果表明,该方法分析精度达到86.2%,对及时掌握工程实施进度及质量,也具有极其重要的意义。
By taking ALOS as remote sensing information source and using supervised classification and human-computer interactive interpretation,the actual area of returning cropland to forestland project is extracted.Since complete measurement for the whole area is expensive and time-consuming,the map-spot area in remote sensing interpretation is selected as auxiliary sampling factor by using sampling methods of probability proportional to size(PPS).The total area of the project is estimated according to the areas of 30 engineering projects and the areas measured by GPS.The areas extracted from GPS and RS are then compared and the precision is up to 86.2%.This study is of importance in controlling the progress of project and its quality.
出处
《水土保持通报》
CSCD
北大核心
2010年第6期22-26,共5页
Bulletin of Soil and Water Conservation
基金
"十一五"国家科技支撑计划课题"国家重点林业工程监测技术研究"(2006BAD23B05)
关键词
遥感解译
不等概抽样
估测精度
RS interpretation
sampling methods of probability proportional to size(PPS)
estimate precision