Desarrollo de un sistema de inspección automático de PCB’S mediante visión artificial

Abstract

The development of the automatic inspection system for PCBs by means of artificial vision makes it easy, simplifies and avoids fatigue problems for the people in charge of carrying out the inspection, of the DINNOVASOFT company. In the initial stage, the hardware is implemented, which integrates the acquisition system through an optical sensor (scanner), and for the digital image processing it is carried out through a computer; For the housing and operation of the hardware of the acquisition system, a metallic structure has been built that contains, a lighting system, in charge of intensifying the light flux over the PCB area, a system that moves the PCB towards the scanner window, by turning a stepper motor controlled by an Arduino, and finally a scanner in charge of capturing and digitizing the PCB image; The next stage is the development of the software, which is carried out on a PC, in Python, OpenCV and PYQT5 programming language, the method is to detect the defect based on the digital image of the PCB, using computational image processing technologies, the algorithm is based on the input of two images, one to be examined for errors and the other is the reference image without errors, these two images when compared and through subtraction, analysis, and recognition operations, resulting in the errors of the PCB, finally these defects can be visualized by a graphical interface.