摘要
本文阐述了GIS和遥感之间的关系,并综合应用GIS信息和遥感光谱数据,进行了改善遥感图象分类精度的试验。选取陕北黄土高原安塞县真武洞为试点,以1987年Landsat-5 TM影象窗原始波段为代表,利用GIS数据库中的地理信息,在以常规光谱特征为基础的模式识别基础上,引进了以光谱效应为基础的模式识别方法及静态层、动态层的概念,通过遥感光谱信息和地理信息分别生成的动态层和静态层两个层理相结合,进行了改善土地利用现状分类精度的尝试,并给出了试验结果。另外对地理信息和影象复合的方法也进行了初步的探索。
In this paper, the relation between GIS and remote sensing was described. ANSAI county, which is located in loess plateau of northern Shanxi, was used as the experimental area and the original multi-spectrum of 1987-landsat 5 TM images was selected. The existing geographical information in GIS data base, was fully utilized. The method of pattern recognization which depends on spectral respondence on the base of pattern recognization by conventional spectral characters, and the concepts of dynamic and static stratification have been introduced The experience was performed to improve the classification of current landuse by combining the static stratification with dynamic stratification. Some results are given in this paper, The research has also been made in. compounding geographical information with images.
出处
《武测科技》
北大核心
1990年第1期37-42,共6页
