Oxidation Detection on PCB Using Image Processing

A Manual visual inspection system in the manufacturing company resulted in decreased inspection performance due to the human error factor. Defects that naturally occurring in the mainboard are oxidation and it cannot be repaired. This study aims to improve the quality of inspection in identifying defects, especially the mainboard which is exposed to oxidation in the area of IC (Integrated Circuit) components. The system uses a digital microscope camera to capture the image of the IC component and using the Microsoft Visual Studio application as the interface. Template matching and the NCC (Normalized Cross-Correlation) algorithm as a method for detecting oxidation. NCC is a calculation of the correlation function between two imageries as a comparison value between the template image and the test object image. The experiment was carried out twice on 10 unoxidized PCB (Printed Circuit Board) samples and 10 oxidation-affected PCB samples. The percentage of successful detection of IC components in this study reached 100% with the accuracy or thoroughness value of 90%. The light intensity greatly affects the percentage outcome due to the inconstant light intensity resulting in a percentage value change from testing.