An image processing application to detect faulty bottle packaging

Use of automation technologies has received significant attention in recent years and has become one of the crucial parts of industry. The inspection of outputs in a production process is of great importance for companies to produce faultless products. This is vital for companies to sustain credibility and reliability as well. In this study, it is aimed to detect faulty packaging of bottles. Mineral water is bottled in factories and packaged to produce six-bottle and twenty-bottle packs transported on a conveyor belt. Faulty packaged products are rarely observed in this production process. However, it is not acceptable for these faulty products to leave the production line. In this work, an image processing application is implemented to detect and label faulty bottle packages for impeccable production. Sobel filter is utilized for edge detection and Hough transform is used for calculation of number of bottles in packages. The ones labelled as faulty packages are eliminated from conveyor belt by means of a boxer motor.