透過您的圖書館登入
IP:34.236.152.203
  • 學位論文

應用快速視覺定位技術於TFT-LCD覆晶薄膜自動化檢測

Applying Image Alignment Technology to Automated Visual Inspection of TFT-LCD COF

指導教授 : 田方治
若您是本文的作者,可授權文章由華藝線上圖書館中協助推廣。

摘要


本論文研究對象為TFT-LCD覆晶薄膜(Chip on Film;COF),由於COF印刷電路精細與微小之特性,COF上的瑕疵肉眼不易察覺。目前工廠以人工目視檢測方式找出瑕疵,但成本過高,且員工容易因長時間工作使眼睛疲勞而導致人為檢測疏失,降低檢測之可靠度,使出貨的品質降低,進而影響公司信譽。 本研究目的以自動化光學檢測方式,取代現行人工目視檢測,以改善人工目視檢測缺點。由於生產線待檢測的COF有旋轉、平移的現象,所以為達自動化檢測方式,影像定位環節相當重要,若沒有良好的定位機制,將造成系統瑕疵誤判率大大提升。本論文提出一種快速區域特徵定位技術,利用物件標記演算法,將影像上的物件區出來,接著利用區域描述子計算物件特徵,再利用最小距離分類器進行物件比對,最後計算物件偏移縮放旋轉角度。本研究之定位技術精確度優於傳統的定位方式,且經標準化影像後,可成功改善環境照度影響定位功能的缺點,定位執行時間與精確度優於套裝軟體eVision的定位技術。 本系統檢測一張COF的平均時間為3.24秒,工廠管理者希望將此AOI系統架設於生產線的電路檢測站,而電路檢測站檢測電路的時間約為4秒,所以,本AOI系統仍然可應付COF全數檢驗之效果。經實驗測試,本研究之TFT-LCD覆晶薄膜自動化檢測系統,正確率達96.5%。

並列摘要


Due to the COF printed circuit is accurate and small, our eyes are not easy to find the defect. The laborer find the defect of COF by their eyes in the factory at present. Although this method can reduce defects of product but it is very expensive and easy to cause eyes to be tired and it would reduce reliability of inspection and worsen the quality. Then the company’s reputation will be worse. Automated Visual Inspection is used to replace human eye’s inspection. Due to COF would shift and rotate on the conveyer, the image alignment is very important in our system. If the inspection system does not have good alignment method, it will cause erroneous judgement. This thesis proposes a fast image alignment algorithm consists of three steps: 1) proposed improve component-labeling, 2) Area descriptor transformation, and 3) orientation estimation. This algorithm is superior to the traditional template matching. We can improve environmental illumination influence with standardized image. The execution time of image alignment is better then package software. The average time of inspecting one piece of COF is 3.24 second,This system can inspect all COFs with high speed and the accuracy still can reach 96.5%.

參考文獻


[36] 陳楙開, (2004) ”以位移暫存器為記憶體模組的區塊比對法之積體電路設計”, 中原大學碩士論文
[39] 葉郡維, “料盤中之液晶顯示器驅動IC瑕疵檢測", 國立台北科技大學自動化科技碩士,2007
[31] 李文昌, " 移動物體即時追蹤研究", 元智大學電機工程研究所碩士論文, 2002
[3] A. K. Jain, R. P. W. Duin, J. Mao, and M. Senior, “Statistical Pattern Recognition: A Review,” IEEE Transactions on pattern analysis and machine intelligence, Vol.22, No.1, 2000, pp.4-37.
[4] B. Zitov´a, J. Flusser, “Image Registration Methods: A Survey,” Image and Vision Computing, vol. 21, no. 11, pp. 977.1000, 2003

延伸閱讀