Automatic Recognition of ArUco Codes in Land Surveying Tasks

Abstract

Photogrammetry is experiencing its renaissance in the 21 century. The main driving forces are research in computer vision as well as drone (UAV/UAS/RPAS) technology. There are popular open-source libraries and applications available in these fields (e.g. OpenCV and ODM). In this paper, we report our experiments using OpenCV in specific land surveying projects. Our efforts concentrate on two areas. The first one is the application of close-range photogrammetry for deformation observations. Fitting a simple camera to the eyepiece of a total station we reached sub millimetre movement detection from few meters. This method was used for dynamic test load of a bridge and a floating platform. The second area is automatic ground control point (GCP) detection in images taken from drones. Optimizing the size and colours of the used ArUco markers we achieved nearly 100 percent results in different light conditions.