The growth of the electronics industry lead to a need for efficient methods of testing and validation of printed circuit boards (PCB). It is necessary to identify defects that might appear in a component in early production stages. This task may be performed by automatic inspection systems, showing advantages in speed, accuracy and repeatability, over human inspection.
This paper describes a visual inspection system that is able to detect solder paste deposition defects on a PCB. The image of the PCB is analysed to segment the areas were the solder paste is present and then, by comparing with reference data from the PCB design files, defects are identified, either because of missing or excess solder. The system is based on low cost components, namely a Raspberry Pi Compute Module and two Raspberry Pi v2 cameras.
Experimental test performed with the prototype regarding the defect detection in PCBs and execution time allow to conclude the system can be used in a production line for human inspection aid.