基于遥感影像中电磁特性的地物提取方法  

On Electromagnetic Characteristics Extracting of Surface Objects Based on Remote Sensing Image

作　　者：陈瑞[1,2] 张德锋[2] 薛原[2] 汪连栋[2] 

机构地区：[1]海军航空工程学院 [2]中国人民解放军63880部队

出　　处：《电光与控制》2009年第1期71-74,共4页Electronics Optics & Control

摘　　要：电磁特性数据是电子信息装备内场仿真试验不可或缺的数据资源,为了克服传统电子信息装备仿真试验中的电磁特性数据获取方法的不足,提出了从遥感影像中大批量获取大区域范围的电磁特性数据的新方法。首先依据不同地物的电磁特性规律对某一地域的地物进行了分类,从多尺度分割概念和基于异质性最小的区域合并算法两个方面,深入研究了针对地物分类的面向对象多尺度分割算法,并结合实践对算法进行了应用。由于遥感影像的真实性,保证了电磁特性数据提取的准确性与可靠性,实践证明此研究为大批量获取电磁特性数据提供了一种有效的方法和思路。The electromagnetic characteristic information of each sort of ground object is necessary for fieldlevel test and shop-level simulation. Since it is difficult to obtain it through traditional method, we put forward a new method for extracting the data of electromagnetic characteristics from the remote sensing image. First, the classification of the electromagnetic characteristics of ground objects are educed on the basis of the radar image characteristics, and the surface objects in a certain area are classified according to the electromagnetic characteristics of the different ground objects. Then, considering the concept of multiscale segmentation and the area combination arithmetic based on the minimum principle of heterogeneity, the object-oriented multi-scale segmentation for ground-object classification is studied, which are applied in practice. The facticity of remote sensing image guarantees the accuracy and reliability of information extracting.

关 键 词：电磁特征 地物特征分类 遥感影像多尺度分割方法 信息提取算法 仿真试验 

分 类 号：V271.4[航空宇航科学与技术—飞行器设计]