从遥感TM影像分类图提取小班界限的算法实现

Algorithm Realization of Extracting Edge of Subcompartm ents by Computer Based on TM Classfication Image

在对吉林省汪清林业局TM影像进行计算机自动分类的基础上 ,设计算法完成分类图像的“孤岛”去除、边界提取和宽边缘的弱化功能 ,实现了计算机代替人进行区域面积、形状的分析判断 ,使计算机能自动搜索到真正的边界 ,为制作林业专题图打好基础。通过与吉林汪清林业局提供的地面调查成果的对比分析表明 ,实验小班在位置和形状上都与地面人工勾绘的结果相吻合。

On the basis of image classification, the computer algor ithms were developed to remove the image isolated islands, extract the image edg e, thin the wide edge,and analyses image area and image shape with the help of c omputer automatically.Finally the real edge pixels were obtained.All these proce sses laid the foundation for producing the thematic map of forestry.The comparis on and analysis were made for the subcompartment edges based on auto-produced i mage and the map obtained from field survey,the feasibility of computer algorith m can be accepted.