A CAD matching method for 3D thermography of complex objects

Quality control of complex components using thermographic cameras remains a challenging task in industry because it is difficult to quantify the defect size and locations. This research describes a new automated methodology with minimal user input which uses a 3D pose estimation technique to map infrared thermographic images directly on a CAD file of the component. This mapping allows direct inspection of the thermal measurements on the CAD file. The methodology is compared with state-of-the-art 3D thermography techniques which use additional hardware like range cameras and 3D scanners to obtain 3D data. In our use cases, the proposed methodology provides a better 3D mapping of the thermal images. Traditional techniques typically rely on multiple sensors with different fields of view. This difference results in a partial overlap of spatial data and thermal data, hence a part of the thermal data is discarded in the 3D mapping. This is in contrast to our methodology which only uses one camera and maps all image-points representing the object to a 3D location.

Jaar van publicatie:2019

Trefwoorden:A1 Journal article

BOF-keylabel:ja

BOF-publication weight:1

CSS-citation score:1

Authors from:Higher Education

Toegankelijkheid:Open