液晶顯示器之塗佈式增亮膜表面瑕疵檢測與分類

Surface Defect Detection and Classification for Coating Brightness Enhancement Film in Liquid Crystal Display

湯文正(Wen-Cheng Tang)

指導教授 ： 葉繼豪

國立臺北科技大學/管理學院/工業工程與管理系所/碩士(2006年)

摘要

本論文針對液晶顯示器中之零組件－塗佈式增亮膜(Brightness Enhancement Film, BEF)進行其表面瑕疵之檢測與分類，首先利用機器視覺(Machine Vision)與影像處理(Image Processing)技術偵測增亮膜之表面瑕疵如明暗不均(Mura, Uneveness)、氣泡(Bubble)、線條(Streak)與異物(Contamination)等四種主要瑕疵。利用區域統計參數方法對原始瑕疵影像進行影像增強以凸顯瑕疵，隨後利用Otsu二值化方法將瑕疵與良好部分分離以獲得瑕疵之位置與形狀。再使用最適橢圓法求得瑕疵部分之幾何特徵如瑕疵群數、瑕疵面積、長軸長度、短軸長度和長短軸之比值。最後使用倒傳遞類神經網路(Back Propagation Network, BPN)、機率神經網路(Probabilistic Neural Network, PNN)和最鄰近鄰居法(K Nearest Neighbor, KNN)分別進行上述四種瑕疵之分類。實驗結果顯示本論文提出之方法可有效地偵測出增亮膜上表面瑕疵(Mura、氣泡、線條和異物)之位置與形狀，並利用最適橢圓法計算出其幾何特徵作為BPN、PNN和KNN之輸入資料，經過適當調整參數後可有效地將上述四種增亮膜表面瑕疵精確分類。

關鍵字

增亮膜 ； 機器視覺 ； 影像增強 ； 最適橢圓法 ； 表面瑕疵檢測 ； 瑕疵分類

並列摘要

This paper develops an automatic optical inspection (AOI) system to inspect the surface defects such as Mura, bubble, streak, and contamination on Coating Brightness Enhancement Film (BEF). At first, using histogram equalization and local statistic parameter to enhance the contours of surface defects. Then, using the Otsu threshold method to segment the area of defect and obtain their locations and shapes of the defects. Once the location and shape of a surface defect is addressed, best fitting ellipse algorithm are utilized to extract the geometric features such as the number of the defective groups, area, length of major axis, length of minor axis, and the ratio calculated by the lengths of the major and minor axes. Finally, the back-propagation neural network, probabilistic neural network, and K-nearest neighbor method are used to classify the defect types. Experimental results have shown that the proposed method is able to achieve 100% identification rate by appropricate classification method and parameters.

並列關鍵字

Brightness Enhancement Film (BEF) ； Machine Vision ； Image Enhancement ； Best Fitting Ellipse ； Surface Defect Detection ； Defect Classification

參考文獻

[6] 黃朝義，「偏光膜介紹」，光電產業與技術情報，第20期，1999，第40-47頁。

[21] 陳志忠，液晶顯示器的像素點缺陷與相對亮度均一性之自動化檢測，碩士論文，中原大學機械工程研究所，中壢，2000。

[10] U. Eisemann, “Influence of Spatial Luminance Interactions on Perceived Brightness”, Proceedings of SPIE, 4300, 2000, pp. 184-190.

[11] E. N. Malamas, E. G. M. Petrakis, M. Zervakis, L. Petitm, and J. D. Legat et. al., "A Survey on Industrial Vision Systems, Applications and Tools", Image and Cision Computing, 21, 2003, pp. 171-188.

[12] L. Lin, T. K. Shia, and C. J. Chiu, “Fabrication and Characterization of IC-Processed Brightness Enhancement Films”, 1997 International Conference on Solid-State Sensors and Actuators, 2, 1997, pp. 1427-1430.