双因子抗差贝叶斯估计及其在GIS数据采集质量控制中的应用  

Bifactor robust Bayesian estimation and its application to quality control of GIS data capture

作　　者：张菊清[1] 

机构地区：[1]长安大学地质工程与测绘工程学院,陕西西安710054

出　　处：《测绘工程》2006年第6期20-23,共4页Engineering of Surveying and Mapping

摘　　要：在对GIS扫描数字化过程中,所采集的数据里经常含有系统误差和异常误差。在纠正这些误差的过程中,应同时考虑控制点坐标和数字化坐标的先验协因数。文中以数字化的系统误差拟合模型为例,提出双因子抗差贝叶斯估计误差拟合方法,并分别就控制点含有粗差、数字化点含有粗差和两者均含有粗差3种情况建立相应的模型,实例验证了其抗差效果。In the process of GIS scan digitizing, there are many errors including systematic errors and gross errors in the obtained data. Accordingly, the prior covariance of control-point coordinates and digital-point coordinates should be considered at the same time. Based on the model of systematic error fitting, a method of bifactor robust Bayesian estimation is proposed , and three models under outlier errors in control-point coordinates, digitalpoint coordinates and both of them are provided respectively.

关 键 词：数字化 粗差 系统误差 双因子抗差估计 

分 类 号：P208[天文地球—地图制图学与地理信息工程]