星载合成孔径雷达遥感及多卫星遥感数据融合方法

SATELLITE SAR REMOTE SENSING AND FUSION OF MULTISOURCE REMOTELY SENSED DATA

合成孔径雷达（ＳＡＲ）数据，以其特有的成像机理和对地物信息的属性描述及其全天时、全天候作业能力，形成了对传统多光谱遥感数据的互补。通过数据融合技术，可以优化遥感信息资源，实现多卫星遥感数据的优势互补，使遥感信息得到最大限度的利用，从而为地学研究中的各种模型方法和决策系统提供科学依据。根据信息的抽象程度不同，融合可分为原始数据融合、目标级融合和决策级融合三个层次。在不同的层次具有相对应的信息融合方法。

In this paper,the authors summarize the fusion methods of remotely sensed data present in internal and external papers.For operating on a special imaging principle and a different way detecting information on surface as well as its capability of mapping in day and night and in any weather condition,satellite based SAR data offer some information that cannot be detected by conventional optical sensors.On the basis of data fusion,the information derived from different sensors can be merged to create a better result for applying study.The fused information provides some academic data for various kinds of modeling methods and decision systems in geoscience research.According to the stages at which the fusions are performed,remote sensing data fusion techniques can be divided into three categories:pixel,feature and decision level based fusions.Different fusion methods are used at different fusion levels.