论文部分内容阅读
合成孔径雷达SAR用于图像匹配导航系统,可大大提高导航精度。然而地形起伏和大气干扰等因素使得SAR实时图存在几何失真,降低了匹配性能。通过对SAR实时图几何失真成因分析,提出了一种二维可分离的几何校正方法,并得出方位向几何失真可跟随成像处理预先校正,图像后处理只需校正距离向。利用数字高程模型DEM和正射图模拟得到了3幅具有不同地形特征的SAR实时图,并分别利用3种几何校正方法对其进行统计实验,得到了不同地形SAR实时图几何校正方法选取的准则。
Synthetic aperture radar SAR for image matching navigation system, can greatly improve the navigation accuracy. However, factors such as terrain fluctuation and atmospheric interference make the SAR real-time graph have geometric distortion and reduce the matching performance. Through the analysis of the cause of geometric distortion of SAR real-time image, a two-dimensional separable geometric correction method is proposed. It is concluded that the azimuthal geometric distortion can be corrected in advance by image processing, and the image processing only needs to correct the distance. Three real-time SAR images with different topographic features were obtained by digital elevation model (DEM) and orthophoto map simulation. Three geometric calibration methods were used to calculate the SAR real-time images respectively. The criteria for geometric correction of real-time SAR images were obtained.