Redes neuronales artificiales para la inspección óptica en el control de calidad de PCB

Abstract

This paper is the result of the research on the ANN (artificial neural network) algorithm applied in decision making in the process of AOI (Automatic Optical Inspection) for quality control for electronic prototyping company, generating models for the assurance of Quality in the PCBs (Printed Circuit Board) product, covering the fields of decision making, quality management, production processes, neural computer systems and artificial vision among others. It is intended to develop an algorithm of ANN that provides an approach to human recognition and perception when performing a quality inspection of the final product, based on image analysis and recognition. We present the theoretical concepts explored and the results obtained. Initially a problem definition was made to model, then the data processing was performed, the artificial neural network model to be applied was selected, then the relevant adjustments were made to the model to finally obtain a simulation and validation of the same by the Histogram method.