Akademik Veri Yönetim Sistemi
23rd IEEE Conference on Signal Processing Algorithms, Architectures, Arrangements, and Applications (SPA), Poznan, Polonya, 18 - 20 Eylül 2019, ss.105-110
Automated inspection of Printed Circuit Boards (PCB) is substantial in decreasing scrap rates and the amount of revision for reliable production. This paper presents a standalone system with a benchmark software for defect detection and classification on bare, assembled boards and solder joints. Bare and assembled board defects are grouped into 4 groups each and solder joint into 2 groups. Focuses were made on gathering the findings in the literature under a compact system along with a comprehensible interface. Additionally, Optical Character Recognition (OCR) engine is integrated to the system to detect written text on integrated circuits (IC) for correct type defect detection. Also, polarity markers are detected via Binary Large Object (BLOB) detection to obtain polarity errors. The hardware used for the inspection is highly cost-effective such that the solely closed environment equipped with machine vision camera and proper illumination is sufficient. In addition, owing to the lightness of the image capturing box, inspection can be held in a diversity of locations. The software of the system and several image pairs to test are available in the repository. It is our hope that the software be used as a benchmark system for the optical inspection of printed circuit boards.