基于GIS和RS的地类划分研究分析

通过高分辨率的卫星影像识别技术及地理信息系统技术同森林资源的二类清查相结合进行地类划分,可大大减轻林业勘察外业工作量,提高分类精度。本研究利用ERDAS IMAGINE8.5遥感图像处理软件对灵寿县的分辨率为2.5 m的SPOT5遥感图像进行几何校正和图像增强预处理,采用监督分类方案的最大似然分类法进行地类划分,其结果分类的精度都超过了80%,各地类所占比例与二类调查资料基本一致。

Division of land type based on RS and GIS

Based on the high resolution satellite Images Identification Technology and Geographic Information System,land types are divided in Secondary Forest Resource Inventory in order to ease the field workload of the forestry survey and improve the classification accuracy.Geometric correction and image enhancement of Spot5 remote sensing image with a spatial resolution of 2.5 m of Lingshou Country was carried out.Then the maximum likelihood classification method was used to divide the land types.The results showed that the classification accuracy was over 75%.The proportion of every kind of the land type was basically in consistence with the Secondary Forest Resource Inventory results.