一种动态分区的边境狭长区域网平差方法

A Method of Dynamic Partitioned Regional Network Adjustment on Narrow National Boundaries

针对内陆边境狭长区域特殊自然环境、气候条件和其他敏感因素引起的区域网平差精度问题,提出一种利用动态分区规划的区域网平差方法。基于高分立体像对影像,利用层次分析法结合不同尺度DEM实现任务区动态分区,将控制信息以最优方式传递至无控区域,并逐步剔除粗差解算出最优区域网平差模型。实验结果表明,该方法生成的边境狭长区DEM高程均方根误差小于7.5m,DOM产品单点定位精度误差小于5m,丘陵与平原区域误差小于1m,相邻影像之间的几何拼接精度优于0.5m。

Aiming at the problem of regional network adjustment accuracy caused by the special natural environment,climatic conditions and other sensitive factors in the inland border narrow strip,a regional network adjustment method based on dynamic partition planning is proposed.Based on the high-resolution stereo image pair,the analytic hierarchy process is combined with different scale DEM to realize dynamic zoning planning,and the control information is transmitted to the uncontrolled area with the optimal scheme,and the optimal regional network adjustment model is gradually eliminated.The test results show that the root mean square error of DEM elevation in the narrow stripe area generated by this method is less than 7.5 meters,the error of single point positioning accuracy of DOM products is less than 5.0 meters,the error of hilly and plain area is less than 1.0 meter,and the geometric mosaic precision between adjacent images is better than 0.5 meter.