The thesis deals with the visual inspection of the surface protection of automotive components. In mass production, an automatic control device is responsible for 100% product control. It consists of industrial robots and a transfer system for product manipulation and imaging systems for capturing an image of the product surface and
analyzing the image to determine the presence of a paint defect. This thesis first presents the operation and management of the control device. Then, we carry out an analysis of the functioning and the correctness of defect identification in the surface protection of the products. We analyzed both images of marked products and defective products. We compiled error statistics and found that the control device is excessively ejecting products, which we call false negative. We analyze the reasons for this and give proposals for changes to error detection settings. We also give proposals for the reworking of mechanical assemblies.