標題: 利用電腦輔助設計資料防止錯打線的視覺偵測系統之設計與開發
A CAD-based Vision Approach for Incorrect Wire Bonding Prevention System Design and Development
作者: 李恕明
Lee, Shu-Ming
彭德保
Perng, Der-Baau
工業工程與管理學系
關鍵字: 視覺檢驗;打線位置檢查;錯打線預防;vision inspection;bonding position check;wrong wire bonding prevention
公開日期: 2008
摘要: 在半導體封裝的製程中,沒有一種自動化的方法能在大量生產環境中正確且有效的檢驗打線位置是否正確。本論文設計與開發了一個能預防錯打線的機器視覺系統,利用影像處理與模擬打線的整合技術,可以在真正打線動作開始前就先檢驗打線位置是否正確,進而避免錯打線的發生。本論文所提出的機器視覺系統不需要先製作樣本供實際的檢驗來判斷打線位置是否正確,如此可節省製作樣本的人工、物料及時間成本。 本論文所提出的機器視覺系統能在半導體封裝廠的大量生產環境中與生產同步進行,不需要搬運產品至其他離線機器進行檢驗,如此可節省搬運的時間。與其他檢驗方法比較,本論文所提出的方法可以有效的解決檢驗中常見的誤偵測與漏偵測的問題,同時本論文也提供了一個利用電腦輔助設計資料來檢驗複雜積體電路零件的新方向。實驗的結果顯示了此一機器視覺系統在大量生產環境中是非常完整健全的,而且是一種能與生產速度匹配的快速檢驗方法。半導體封裝廠目前因為錯打線而產生的不良品率約為 2000~3500 PPM [1] ,可經由本論文的方法完全改善。
No available method can automatically verify the correctness of the wire bonding position in the mass production environment. A novel machine vision based wrong wire bonding prevention (WWBP) system is proposed that can prevent wrong wire bonding prior to actual wire bonding process execution. The proposed approach, which integrates image processing and wire bonding simulation techniques, does not need to bond actual samples for testing and can save the manual effort, material cost and product time. The proposed WWBP system can also be used to work synchronously with the real wire bonding process in the mass production environment so that the material transfer process will be reduced. This approach can efficiently solve the mal-detection and lost detection problems that may occur in other available methods. It also provides a new direction for applying CAD-based vision techniques for complicated IC parts inspection. The experimental results showed that the proposed WWBP system is robust and fast enough to work synchronously with the wire bonding process. With the proposed WWBP system, wrong wire bonding incidents, which are about 2000~3500 PPM [1] in current IC packaging foundries, will be totally eliminated.
URI: http://140.113.39.130/cdrfb3/record/nctu/#GT008933801
http://hdl.handle.net/11536/78879
顯示於類別:畢業論文


文件中的檔案:

  1. 380101.pdf

若為 zip 檔案,請下載檔案解壓縮後,用瀏覽器開啟資料夾中的 index.html 瀏覽全文。