摘要
基于距离多普勒模型的SAR图像系统级几何校正存在一定的误差,需要引入地面控制点来对模型参数进行修正来提高校正精度。针对人工选取地面控制点工作量大且模型参数解算复杂的问题,提出矢量地图作为控制信息,利用SAR图像与矢量地图自动配准的方式实现SAR图像的几何精校正。该方法避免几何精度校正中繁琐的控制点选取过程以及复杂的模型参数解算过程。结果表明,该方法的平均相对精度达到10m左右。
There is a certain error in SAR images geometric correction by R D model, so ground control points are introduced to modify the model parameters to achieve a higher correction accuracy. In this paper, the vector map is selected as control information by using SAR image and vector map to automatically register to achieve the geometric correction of SAR image. The result shows that this method avoids the cumbersome selection of ground control points and complex problem of model parameter solving in the process of correction in high precision. After checking, the average accuracy of the relative vector map is about 10m, which is equivalent to the method of manually selecting the control point to improve the accuracy of the SAR image.
作者
吴戌
叶伟
劳国超
WU Xu;YE Wei;LAO Guochao(Department of Graduate Management,Academy of Equipment,Beijing 101416,China;Department of Information Equipment,Academy of Equipment,Beijing 101416,China)
出处
《测绘工程》
CSCD
2018年第10期52-57,63,共7页
Engineering of Surveying and Mapping
关键词
SAR图像
几何精校正
矢量地图
图像配准
道路线特征提取
最优化模型
优化求解算法
SAR image
geometric correction with high precision
vector map
image registration
road feature extraction
optimal model
optimal solution