Analysis the Effect of Control Points Number and their Distribution Patterns on 3D Model Accuracy in Close-Range Photogrammetry Applications

Abstract

The accuracy of products generated by Structure from Motion (SfM) algorithms is influenced by various factors, including the ratio of photo overlap, image scale, the number of control points, their spatial distribution, the resolution of captured images, internal camera calibration parameters, and other related elements. This study aims to evaluate the impact of the number and spatial distribution patterns of control points on the accuracy of three-dimensional models produced through close-range photogrammetry. Control points serve as essential reference markers in photogrammetric workflows, playing a significant role in enhancing the precision of the final 3D model.

The research investigates several configurations that differ in the number and spatial arrangement of control points to determine how these variables affect model accuracy. Through comprehensive analysis of the resulting 3D models, the study identifies optimal strategies for the placement of control points to improve accuracy. The results indicate that the spatial distribution of control points has a substantial impact on the accuracy of 3D models. Among the tested patterns, edge distribution proved to be the most effective in achieving high accuracy with fewer points, while a combined central-edge distribution demonstrated consistent and reliable performance across different point quantities. In contrast, although random distribution showed strong initial results, its performance became unstable as the number of control points increased.