An Experimental Design Approach on Georeferencing

Abstract

Georeferencing is one of the most important stages of digitizing analogue maps. It is affected by many factors such as; scales and resolutions of maps, the number of control points, etc. In this study, four of these factors were investigated using 2(4) factorial design in two dimensional georeferencing of cadastral maps. Factorial design determines, whether the selected factors have main and/or interaction effects on a response variable or not. Map scale, resolution of raster map, the number of control points and the coordinate transformation method were selected as experimental factors. Then, main effects and interactions between these factors were investigated. The results were statistically analyzed using analysis of variance (ANOVA), and a regression model was suggested to consider the significant main and interaction effects of factors. It was observed that the two dimensional georeferencing of maps were affected by each of the selected experimental factors and by the interaction between the map scale and coordinate transformation method.