A Full-Factorial Design Approach for Coordinate Transformation

Abstract

The aim of this study is to analyze the role of experimental factors in the affine coordinate transformation of cadastral maps using a 2(3) factorial design. The selected experimental factors and their respective levels include a map scale of 1/500-1/1,000, a number of common points 4 and 20, and the least square and least absolute value adjustment method. The results were statistically analyzed using Student's t test, analysis of variance (ANOVA), and an F test to define the important experimental factors and their levels. A regression model that considers the significant main and interaction effects is suggested. It was determined that the affine coordinate transformation of cadastral maps was affected by all of the selected experimental factors and the interaction effects between the map scale and number of common points. As results of the experiments, it is determined that the factors levels for the reduced response variable are the 1/500 map scale, 4 common points and the Least Square Adjustment method.