人工智慧化瑕疵電路標記系統

A Circuit Defect Marking System with Artificial Intelligence

連健宏 ； 林義暐 ； 王裕銘 ； 李偉杰 ； 梁隆禎

《機械工業雜誌》 438期 (2019/09) Pp. 16-23

摘要

近年來，為因應軟性電路板的檢測市場需求及技術變革，工研院機械所致力於人工智慧化電路瑕疵打標系統的開發，以提升國內軟性電路板業者及光學檢測業者的技術能力及產品價值。人工智慧化電路瑕疵打標系統搭配機器學習缺陷辨識技術、瑕疵定位串接技術及機器手臂打標技術，針對細線寬軟性電路板的檢測，可有效降低人力成本、減少產品誤檢率及提高產品良率。該項技術除可應用於軟性印刷電路板外，也可用於面板及半導體…等線路的檢測。本文針對人工智慧化電路瑕疵打標系統及其設備進行深入探討。

關鍵字

自動光學檢測 ； 人工智慧 ； 瑕疵打標系統

並列摘要

In order to meet market demand and to advance technical capability in flexible printed circuit board (FPC), MMSL (Mechanical and Mechatronics Systems Research Laboratories) develops a circuit defect marking system with artificial intelligence to improve domestic FPC quality and to add product values. The artificial intelligent defect marking system involves various technologies, such as defect identification technology, defect positioning technology, and laser marking with robotic arm. The system can effectively reduce labor cost, minimize inspection errors, and improve yield. The technology can also be applied in other advanced electronic devices such as flat panel display and semiconductor industry, etc. In this article, the artificial intelligent defect marking system for fine-line flexible printed circuit board will be reviewed and discussed in depth.

並列關鍵字

Automatic optical inspection ； Artificial intelligence ； Defect marking system