【摘要】 合成孔径雷达(SAR)图像特殊的斜距成像方式使得利用传统的校正方法很难校正雷达图像。常用的方法是利用DEM和SAR成像几何信息模拟的SAR图像对真实SAR图像进行校正。在区域制图中,平坦地区由于没有地形起伏特征,使得模拟SAR图像特征不明显,无法建立与真实SAR图像的配准关系。本文在研究大范围SAR图像自动化地理编码的过程中,分析讨论了图像模拟和图像匹配的原理和过程,实现了自动判断模拟SAR图像与真实SAR图像的匹配偏移量是否达到信噪比要求,从而自动判断是否使用具有地物专题信息的TM图像代替模拟SAR图像和真实SAR图像进行匹配,实现了海量SAR图像地理编码的自动化。更多还原

【Abstract】 The special slant-range imaging mode of Radar image makes it difficult to use traditional correction method to geocode it.Commonly we simulate a radar image using DEM and SAR geometry information to geocode a real SAR image.In the application of regional cartographic imaging,when we encounter flat areas,the features of simulated SAR image are not so obvious,making its failure to match the real SAR image and failure of automatic correction of a wide range of SAR image.In this paper,we analyzed and discussed the principle and process of the image simulation and image matching.By achieving the automatic determination of the offset between radar image and simulated SAR image matching the signal-noise ratio requirements or not,we can automatically determine whether or not to use the TM image which contain thematic information of feature,instead of the simulated SAR image for matching with the real radar image.The method achieves automatic geocoding of mass SAR image as the same time.更多还原