基于总体最小二乘的相似变换模型及其在地图扫描数字化中的应用

Application of similarity transformation model based on total least-squares to map scanning digitization

在地图扫描数字化中,图纸定向的精度直接影响后续地图要素数字化的精度,而在图纸定向中较为常用的相似变换模型未能周全考虑地图定向点的采集误差以及原始地理坐标的测量误差。引进总体最小二乘方法,建立基于总体最小二乘的相似变换模型,结合地图扫描数字化实验,分析基于总体最小二乘方法的相似变换模型的适用性及优越性。

For map scanning digitization,the accuracy of drawing orientation directly affects the accuracy of follow-up digitization of map elements,but in drawing orientation,the common-used similarity transformation model fails to consider the acquisition errors of the map directional points and the measurement errors of original geographic coordinates.The total least-squares method is introduced to set up a similarity transformation model according to total least-squares.Combined with map scanning digitization experiment,analysis is made on the applicability and advantages of the similarity transformation model based on total least-squares method.