臺灣博碩士論文加值系統

現今工業領域的自動化製程越來越以提高產品的生產效率以及精密度為主，而自動化光學檢測(Automatic Optical Inspection, AOI)扮演著快速及精確檢驗的關鍵技術，不僅取代大量耗時人工檢測的生產成本，更可大幅的提升生產的產能與品質，但是自動化光學檢測系統對程式編寫與檢測相機要求非常高，一般都需以工業用相機與誠實語言來做檢測與編寫。
本研究檢測產品為印刷電路板(PCB板)，其PCB板加工過程中容易產生錫膏漏印、錫膏短路、擴孔等缺陷，在現今的設備下多是利用工業攝相機做檢測並利用程式語言軟體做軟體設計，檢測物件非常單一，本論文在不使用工業相機與程式語言設計的情況下、利用一般網路攝影機以垂直面對PCB板上半層之區域取得影像，再做機械視覺缺陷檢測，並運用三菱的可程式邏輯控制器(PLC)作為檢測機台之自動化模組，配合NI Vision Builder AI 為影像處理與檢測模組，將其整合，進而達到自動化PCB板缺陷檢測之目的。最後根據不同缺陷樣貌、環境條件及影像處理方法進行缺陷辨識檢測。
本論文實驗在驗錫膏漏印、錫膏短路、擴孔等缺陷上可以零誤判，一張影像所需要演算的時間為323ms，並能檢測複數以上的物件，所要之設備費用也不到一般市面上光學檢測設備的五分之一之，因為使用PLC做為控制也方便各家廠商銜接至現有的生產線上，在影像辦判斷精度的部分也可以依各家廠商規範做調整不需經由專業的程式語言編寫者做修改，進而可以省掉大量後期設備調教保養之費用。

At the present industrial sector, the automation engineering solutions have become more and more popular to improve the productivity and precision for the manufactured goods. The Automatic Optical Inspection (AOI) plays the role of a key technique for a rapid and precise inspection. Such technology has replaced the labor cost for manual inspection as well as increase production capacity and quality. However, AOI’s requirement for camera and computer programming are very high. In general, an industrial camera and programing language are required for inspection and editing.

The research product for this study is the printed circuit board (PCB). The common defects during machining process for PCB include solder paste leakage, solder paste short circuit, and hole expansion, etc. Nowadays, most facilities use industrial standard camera for detection and programming language for software development. The detected objects are very simple. This study is performed under the conditions of not using industrial standard camera nor programming language but only uses standard webcam that’s held in a vertical position facing PCB’s upper sector to capture an image and then perform Machine Vision Default Inspection along with Mitsubishi’s programmable logic controller (PLC) as the automation module of the inspection device; coordinate with NI Vision Builder AI for image processing and inspection module, then integrate all to achieve automated PCB defect detection. Lastly, according to different defect appearances, condition of the circumstances and image processing methods to perform defect inspection.

In this study, the purpose of this experiment is to get an absolute correct result on defect inspection; which include solder paste leakage, solder paste short-circuit, and hole expansion. An image requires 323 minutes to determine, and it can detect multiple objects at a one- fifth of the cost compared to optical inspection equipment on the market. Because the use of PLC as a controller, it is also convenient for the manufacturer to connect to the existing production line, the accuracy of the image detection can be adjust according to the specification of the manufacturer without the need for programmer to make changes, which saves a lot on the equipment adjustment and maintenance costs in long term

摘要...................................i
Abstract..............................ii
誌謝..................................iv
目錄...................................v
表目錄.................................vii
圖目錄................................viii
第一章、緒論............................1
1.1前言................................1
1.2研究動機與目的.......................3
1.3研究方法與步驟.......................3
1.4論文總覽.............................5
第二章、文獻回顧.........................6
2.1數位影像基礎概論......................6
2.1.1影像二值化..........................7
2.1.2影像濾波............................8
2.1.3特徵邊緣輪廓.........................9
2.2影像檢測之應用文獻回顧.................10
2.3 PCB板檢測文獻回顧.....................14
第三章、實驗設置..........................16
3.1 PCB缺陷之特徵........................16
3.2自動化影像檢測機台設置.................18
3.2.1機台設計與組裝......................18
3.2.2影像擷取單元........................20
3.2.3自動化控制單元......................21
3.2.4驅動單元...........................22
3.2.5 外部電源單元.......................22
3.2.6 感知單元...........................23
3.2.7 配線...............................24
3.3各缺陷檢測方法及步驟....................25
3.4自動化整合及設計........................32
3.4.1自動化動作流程及設計...................32
3.4.2自動化影像檢測機整合方法及步驟..........34
第四章、實驗結果與討論.....................38
4.1機構軟體設置可靠度分析(標準尺寸分析)......38
4.2 PCB缺陷初步檢測結果....................41
4.2.1良品PCB板............................41
4.2.2錫膏漏印PCB板........................42
4.2.3錫膏覆蓋區域不當PCB板.................42
4.2.4擴孔PCB板............................43
4.3檢測參數分析............................44
4.3.1檢測參數校正..........................44
4.3.2 50組檢測檢測結果.....................45
4.3.3機電整合..............................46
4.4系統精度探討............................47
4.5與工業用攝相機檢測系統差異討論............51
第五章、結論與未來展望......................52
5.1結論....................................52
5.2未來發展方向.............................52
參考文獻....................................53
附錄A-1.....................................55
附錄A-1.....................................57
附錄B.......................................59
附錄C.......................................61
附錄D.......................................64
Extended Abstract...........................67

[1]張文陽，2015，機器視覺與影像處理課程之教材，國立虎尾科技大學。
[2]白楚平，2017，” 以LabVIEW系統實現即時車牌偵測與辨識”，國立台灣海洋大學機電與機械工程研究所碩士論文。
[3]陳俊宏，2011，”線上型葉綠素螢光影像檢測蘋果瘀傷系統之研製”， 國立宜蘭大學生物機電工程研究所碩士論文。
[4]溫梓宏等，2004，” 整合機器視覺於自動化油封修邊機之線上尺寸檢測製程”，建國科技大學自動化工程系暨機電光系統研究所碩士論文。
[5]譚孟桓，2014，”機械視覺應用於四軸機械手臂之定向夾取”，龍華科技大學機械工程研究所碩士論文。
[6]陳凱傑，2012，”柴油引擎共軌系統回流管套光學檢測系統之開發”，國立台灣科技大學機械工程研究所碩士論文。
[7]林姍昀，2013，”利用LabVIEW於自動追蹤及定位系統支架構”，逢甲科技大學航太與系統學系研究所碩士論文。
[8]李紹誠，2009，”應用LabVIEW虛擬儀控於雷射動態追蹤系統之建構”，逢甲科技大學航太與系統工程研究所碩士論文。
[9]黃琨展，2016，”自動化玻璃面板檢測系統人機介面設計”，崑山科技大學機電工程研究所碩士論文。
[10]王偉杰，2015，”微精密鋼珠表面瑕疵檢測系統開發”，國立台北科技大學工業工程與管理研究所碩士論文。
[11]蔡佳融，2016，”汽缸孔內壁缺陷之自動化影像檢測機台研究與開發”，國立虎尾科技大學機械與電腦輔助工程研究所碩士論文。
[12] Shanmugamani R., Sadique M.and Ramamoorthy B., 2015, "Detection and classification of surface defects of gun barrels using computer vision and machine learning", Measurement, 60, pp. 222–230
[13] Khan M.F., Khan E. and Abbasi Z. A., 2015, "Image contrast enhancement using normalized histogram equalization", Optik, 126,pp. 4868–4875.
[14] Wu Z., Lu X. and Deng Y., 2015, "Image edge detection based on local dimension: A complex networks approach", Physic A, 440, pp. 9–18
[15] Rong Z., Li Z. and Li D. N., 2015, "Study of color heritage image enhancement algorithms based on histogram equalization", Optik, 126, pp. 5665–5667
[16] Wu W. Y., Wang M.J. and Liu C. M., 1996, "Automated inspection of printed circuit boards through machine vision", Computers in Industry, 28, pp. 103- 111
[17] Shen H., Li S., Gu D. and Chang H. X., 2012, "Bearing defect inspection based on machine vision", Measurement, 45, pp.719–733
[18]郭家豐，2018，”印刷電路板導電區域瑕疵自動檢測系統之研究”國立高雄大學資訊工程學系研究所碩士論文。
[19]林雍勛，2008，”X射線層析攝影合成法之PCB板缺陷檢測”國立清華大學動力機械工程學系研究所碩士論文。
[20]喻振庭，2014，”改良型Qtsu法在PCB自動光學檢測系統之應用 ” 國立台灣科技大學電機工程學系研究所碩士論文。
[21]林宸生，2014，”應用自動化光學檢測於印刷電路板鑽孔製程檢測之研究” 逢甲大學資訊電機工程學系研究所碩士論文。
[22]劉權霈，2014，”應用電腦視覺技術於PCB自動檢測系統之設計開發。
[23]F. R. Leta, F.Feliciano, “Computational System to Detect Defects in Mounted and Bare PCB Based on Connectivity and Image Correlation,”Systems， Signals and Image Processing，pp. 331-334, 2008.