Geometric Calibration of a Camera System of an Optical Permeameter

Industrial camera systems are essential for performing optical measurements in both research and manufacturing. However, perspective and lens distortions may cause unwanted errors, which compromise measurement accuracy. Geometric camera calibration provides a method to mitigate these issues. Additionally, camera calibration enables to extract measures in metric units from dedicated camera images. This thesis focuses on the geometric camera calibration of an optical permeameter's camera system. This test rig is used in the field of fibre-reinforced polymer composites. With the optical permeameter, radial flow experiments are conducted to determine the in-plane permeability of textile reinforcements. The textile fabric's in-plane permeability is determined by optical tracking of the flow front advancement, which is captured through a series of images using an industrial camera system. This material property is important for flow simulation in liquid composite moulding and thus plays a substantial role for the optimisation of process parameters. The primary objective of this master thesis is to investigate whether geometric camera calibration can improve the measurement accuracy of an optical permeameter. Therefore, a geometric camera calibration routine is developed for the camera system of the optical permeameter at the Chair of Processing of Composites and Design for Recycling at Montanuniversität Leoben. A comprehensive literature review establishes the mathematical foundations of geometric camera calibration, including projective geometry. The intrinsic and extrinsic parameters, specifying the camera calibration, are introduced, along with different models of lens distortion. A comparative study of different camera calibration routines is conducted. Zhang's camera calibration routine is identified as the most suitable approach regarding the requirements given by the permeability test rig. This routine is integrated into the optical permeameter's existing control software, which is implemented in LabVIEW. The new implementation is evaluated through comparisons with the existing version, which relies on a fixed scaling factor. To achieve this, the deviation of measured distances from known real-world distances is calculated for representative reference images. For these reference images, the implemented geometric camera calibration routine leads to more accurate distance measurements in the image than the method based on a fixed scaling factor. The results demonstrate that the implemented geometric camera calibration routine can improve measurement accuracy in determining textile in-plane permeability using optical permeameters. This work contributes to the development of camera calibration methods for optical measurement systems and highlights the potential of geometric camera calibration to enhance the precision of optical measurements.