Studying the Effect of the Numbers and Distributions Patterns of Control Points on the Geometric Accuracy of the Orthophotos - Application on an Archeological facade

Abstract

Ground Control Points (GCPs) are used for indirect georeferencing in close range photogrammetry systems by phototriangulation (bundle adjustment). This study aims to determine the optimum number and distribution of control points to produce the most accurate orthophoto for a historical façade. In the study, we applied three scenarios for the distribution of control points, namely: border, central, and uniform distribution, starting with four points and then increasing by four points each time. A total of 95 control points and 30 check points were used, and 11 orthophotos were generated in each scenario.               The results showed that the central distribution of control points is the worst among the tested distributions, as the horizontal mean square error values ​​of the orthophoto ranged from 18.5 to 38 mm despite the increase in the number of points from 4 to 35 points. The horizontal mean squared error values ​​ranged between 21 and 11.2 mm in the border distribution and between 34.8 and 6.8 mm in the uniform one. As a result of the tests, it was found that the uniform distribution of the control points, which ensures that there is a point in every 10 m² of the facade, is the best solution to obtain an accurate phototriangulation and accurate orthophoto.