基于QuickBird影像退耕还林工程监测技术研究  

Monitoring the Conversion of Farmland to Forests Project Based on QuickBird Image

作　　者：赵峰[1] 张怀清[1] 鞠洪波[1] 黄建文[1] 

机构地区：[1]中国林业科学研究院资源信息研究所,北京100091

出　　处：《林业科学研究》2008年第z1期1-5,共5页Forest Research

基　　金：科技部社会公益研究专项"国家重点林业生态工程监测与评价"(2003DLA6N014)

摘　　要：采用面向对象的分类方法,充分利用高分辨率遥感影像丰富的纹理信息,选用QuickBird高分辨率遥感影像对退耕还林工程区造林信息进行提取。分类方法采用基于对象间距离的最邻近分类器法和基于规则模型的隶属度函数法。研究结果表明用高空间分辨率的QuickBird影像监测退耕还林工程地理区划、造林施工面积、整地方式及规格等是可行的,造林地块信息提取的准确度和稳定度均在80%以上。Based on the abundant textures of the high resolution remote sensing image,an object-oriented classification was used to extract the artificial afforestation from QuickBird image in the area of the conversion of cropland to forest project.The classification adopted the object-distance-baged nearest classifier method and project model-oriented membership function method.The results indicated that it was feasible to get the layout,area,planting manner and standard of the conversion of farmland to forest project from high resolution remote sensing image.The veracity and stability were more than 88 percent.

关 键 词：高分辨率遥感影像 QUICKBIRD 退耕还林 面向对象分析 

分 类 号：S771.8[农业科学—森林工程]