Visual inspection of coatings in hydrogen fuel cells

In industrial production, the presence or absence of defects on the surface of industrial products will have a great impact on the quality of some products. In this paper, we propose a method that combines traditional image recognition and deep learning-based object detection algorithms, which aims to detect whether there are defects on the coating of hydrogen fuel cells and identify the types and sizes of defects. This method integrates multiple functions for detecting different defects, such as the overall detection and identification of defects such as large area missing coating, uneven application, empty bands and dimensions that do not meet the standard, and detailed detection such as the detection of light transmission points and the detection of light-sladen areas of the coating material (referred to as dark layers in this paper). The system can be installed directly on the production line to detect defects in the coating material of hydrogen fuel cells.