摘要
目前对于无地面控制点的R-D构像方程的求解方法多为采用繁杂的坐标参数变换后进行解析处理。为使求解方法简化,首先将R-D构像方程中的星地距离矢量作为变元进行泰勒级数展开导出误差方程,然后利用最小二乘平差原理迭代求解目标点的位置信息。利用苏州地区ERS-2SAR图像产品和该地区1:10万地形图对该方法进行了实验验证,结果表明该方法平面定位精度达到200m量级,可以满足无控制点条件下中分辨率星载SAR图像中目标像元定位的需求。
Directly locating pixel in SAR image is based on Range-Doppler model without tie points. In general, transforms among different coordinates used to solve Range-Doppler equations, are complex and depend on many projection parameters. In this paper, a new method is developed to solve R-D equations. Firstly, the R-D equations are expanded by Taylor expansion while the distance between SAR satellite and target is viewed as a variable parameter. Then the absolute location is retrieved iteratively by least-square adjustment for linear R-D equations. Test is conducted using ERS-2 SAR imagery of Suzhou area, China, comparing with topographic map with a scale of 1 ： 100,000. The results indicated an accuracy of 200 meters is attainable independent of any reference or GCP point.
出处
《电波科学学报》
EI
CSCD
北大核心
2006年第1期1-5,共5页
Chinese Journal of Radio Science
基金
国家重点基础研究发展规划项目(2001CB309406)
国家自然科学基金项目(40501044)
关键词
SAR
定位
R-D模型
最小二乘
SAR, target location, Range-Doppler model, least square method
