Investigation of Robot-Assisted Photogrammetry with a Moving Camera

Scope

Within the Cluster of Excellence The Martian Mindset – A Scarcity-Driven Engineering Paradigm, production and monitoring systems are investigated for scenarios with strongly limited resources. In the subproject COMPOSE, a robot-assisted measurement system is being developed that aims to perform different geometric measurements with a minimal number of optical sensors. In optical 3D metrology, stereo camera systems are commonly used to reconstruct three-dimensional structures from two cameras with known relative geometry. An alternative approach is to replace one of the cameras by moving a single camera mounted on a robot, thereby creating the stereo baseline through known robot poses. The objective of this thesis is to experimentally investigate how robotic positioning uncertainties affect photogrammetric 3D reconstruction and to what extent a moving camera can replace a fixed stereo baseline.

Possible Tasks

• Literature review on photogrammetric 3D reconstruction and stereo triangulation

• Setup of a simple stereo reference system with two cameras

• Implementation of a reconstruction pipeline (e.g., using OpenCV)

• Development of a measurement setup with a robot-mounted moving camera

• Experimental measurements with reference objects of known geometry

• Comparison of the reconstruction accuracy of both approaches