Quality control of faults in glass products by imageprocessing and deep learning methods

Quality control in the production of glass products, which are frequently preferred in the food sector, is traditionally carried out by expert workers by visual inspection. Since this process progresses based on people, the margin of error has been inevitable. An integration to be made in the first stage of production will ease the quality control process in other stages and reduce the margin of error. In this study, it is aimed to control the quality of glass products by using a computer-aided system using various image processing techniques. Pixel-based image segmentation, linear regression, multi-layer neural network, machine learning are used together. Separation and testing of defects on glass products supplied by Baştürk Cam company were carried out. The correct classification rates of the proposed methods were found to be satisfactory and suggestions were made for future studies.