地图扫描矢量化误差的最小二乘配置法处理研究  

Processing of vectorization error to scanned map based on least squares collocation method

作　　者：岳东杰[1] 梅红[1] 

机构地区：[1]河海大学土木工程学院,南京210098

出　　处：《测绘科学》2007年第2期51-53,共3页Science of Surveying and Mapping

基　　金：江苏省交通科学计划研究项目(05Y01)

摘　　要：屏幕扫描数字化是地图数字化的常用方法。在数字化过程中,由于地图本身的变形、扫描仪分辨率等系统误差会影响数字化的结果。仿射变换是实现栅格数据向矢量化数据转换的常用方法之一,研究表明基于最小二乘的仿射变换只可以消除整体变形,而不能消除局部变形。为此作者提出采用最小二乘配置。采用最小二乘配置可以在计算转换参数的同时估计局部系统变形信号,可以消除或减弱以上误差的影响,改善数字化精度。实例分析表明,是提高扫描数字化的有效方法。Digitizing based on screen is a common way of map-digitizing.In the procedure,some systematic errors such as map distortion and resolution of scanner will influence the digitizing result.The affine transformation is a common method that realizes the transformation from raster data after scanning to vector data.It is found that the transformation only based on least squares can only eliminate the entire deformation,but the local deformation of map and resolution of scanner in the processing arent eliminated.The author commends to use least squares collocation.By the method,we can calculate adjustment parameter and evaluate the influence of local signals at the same time,can eliminate or diminish the influence of these errors,then the precision will be improved,through some study and analysis of a instance,we conclude that the least squares collocation is an efficiency method to advancing the precision of scan digitizing.

关 键 词：地图数字化 最小二乘配置 仿射变换扫描数字化 

分 类 号：P282[天文地球—地图制图学与地理信息工程]