Proposals for Development of the Prospective System for Optical Quality Control of the Assembly of Microelectronic Devices.

Autor:	Balizh, K. S., Eremeev, P. M., Simakhina, E. A.
Předmět:	ARTIFICIAL neural networks OPTICAL control MACHINE learning QUALITY control PRINTED circuits CONVOLUTIONAL neural networks
Zdroj:	Russian Microelectronics; 2023 Suppl 1, Vol. 52, pS246-S250, 5p
Abstrakt:	The solution of tasks of detection and classification of defects of printed circuit boards using machine learning is discussed in this article. Six classes of defects are defined. We carried out manual annotation of defects on images of printed circuit boards in Roboflow. The YOLO v.8 neural network model was used with the settings of its training parameters. To improve the accuracy of defect detection we have expanded the dataset and have conducted the experiments. The results of the automatic search for defects in printed circuit boards are presented and explained. [ABSTRACT FROM AUTHOR]
Databáze:	Complementary Index
Externí odkaz:	Zobrazit plný text záznamu Full text from SpringerLink