軟式隱形眼鏡瑕疵檢測系統之研究

Abstract

軟式隱形眼鏡的發明，帶來了優於硬式隱形眼鏡的方便性外，同時能精準的校正視力，製作時間也比硬式隱形眼鏡快。然而在製造過程中，製造機台在隱形眼鏡成形，或修整隱形眼鏡邊緣，甚至在人工夾取鏡片時都有可能損害鏡片造成瑕疵。這些瑕疵的種類包括了隱形眼鏡內的破洞、傷痕與附加物；外部邊緣的不平整、裂縫，以及人工夾持造成鏡片邊緣破損等。 隱形眼鏡製造目前仍採用人工目視檢查瑕疵，然而製造商希望能大幅提升檢測的效率與正確率。因此，本論文擬研發一隱形眼鏡瑕疵之自動光學檢測機台，針對隱形眼鏡之品質分為二個主要項目進行檢測，分別為鏡片邊緣檢測與鏡片內部表面檢測。 本論文之研究方法主要分為硬體與軟體兩大部分。硬體部分，利用固定焦距鏡頭與CCD搭配合適的光源系統來擷取待測隱形眼鏡影像。方法上主要著重在瑕疵檢測演算法的設計，透過所研發之演算法檢測所擷取的隱形眼鏡影像是否有瑕疵。隱形眼鏡的五種瑕疵進行檢測演算法設計，包括破片、邊不良、破洞、傷痕與附加物等，本論文期望設計ㄧ檢測機台與方法可以找出隱形眼鏡所可能會發生的瑕疵並辨別其種類，以取代目前人工目視檢測作業，提昇檢測速度與降低誤判率。

With the invention of the soft contact lenses, it not only becomes more convenient and takes less time to produce than hard contact lenses, but also corrects sight precisely. During the process of manufacturing soft contact lenses, flaws which include cuts, edges, scratches, tears, and chips may be generated due to inappropriate operation. At present, artificial visual inspection is the most popular way to detect flaws; however, it results in low efficiency and accuracy. Thus, the main objective of this study is to develop an automatic optical inspection (AOI) machine for contact lenses defect inspection. An illumination system with fixed focal lens and charge coupled device (CCD) is used to capture the images of contact lenses. After images are captured, an algorithm is employed to check if there are flaws showed on the images. Five kinds of flaws (cuts, edges, scratches, tears, and chips) can be detected by the designed algorithm. This study expects to develop an inspection machine and to design an algorithm which can detect and identify flaws of soft contact lenses. Therefore, the artificial visual inspection can be replaced by an AOI machine which is able to raise the speed of inspection and lower the probability of misjudgment.