| Περίληψη: | Quality assessment in laser welding is of outmost importance in today’s manufacturing. A plethora of in-line inspection techniques have been developed identifying molten pool geometry and weld defects for quality evaluation. This thesis aims to introduce a quality assessment method for the classification of the welds into different categories. The study corresponds to camera-based monitoring approaches, taking as input dummy images of the thermal field captured during laser welding. At the first stage, an image processing technique for the identification of weld defect’s location and size will be deployed. Moreover, machine learning techniques will be also considered for the classification model development and the prediction of the quality status. Finally, requirements specifications and the design of a cloud based application for real-time monitoring and quality indications of the welds will be also described within the thesis
|
|---|
