微型銑刀自動化光學量測與檢測系統之研究與開發

Abstract

PCB 生產製程中，由於產能需求大、精密度要求高，再加上板材複雜化，因此高轉速、高精度、高進刀速的成型機已漸被採用，而要實現高速成型及高精度加工目的，PCB成型銑刀的品質便是關鍵因素之一。 銑刀的外徑、刃長與刀刃的完整性，為其品質評估的指標。傳統上，微型銑刀的外徑，量測多數是利用雷射量測儀來進行。然而雷射量測仍有定位上的問題，而微型銑刀其刀刃上的瑕疵，傳統上多利用高倍率的顯微鏡放大後，再經由人工進行檢驗，此方法也會有因目視檢測而造成檢測品質不穩定的問題。 因此，本研究旨在開發出以機器視覺技術為基礎之微型銑刀的量測檢測系統，此系統可以用來量測微型銑刀的刀徑、刃長與檢測微型銑刀常見的刀刃上的瑕疵，如缺角、崩刃、魚尾缺損與無魚尾等瑕疵。且希望此系統之準確度可以作為維持微型銑刀良好品質的輔助工具。然而微型銑刀有許多不同的種類，其應用範圍也不盡相同，所以在本計畫中，微型鑽石目銑刀是檢測的對象。由於鑽石目銑刀有許多不同的規格，所以量測刀徑與刃長是用來確保微型銑刀基本品質的重要指標。然而檢測刀徑與刃長僅可確保微型銑刀的基本品質，因此仍須更進一步檢測其常見的刀刃缺陷，要檢測這些缺陷，即需要有特殊設計的光源輔助照明，用來凸顯出刃面紋路的特徵，方可取得其清晰之影像，再利用影像處理的演算法找出缺陷的所在，以更進一步確保微型銑刀的品質。本研究擬採用最小灰階值影像當作微型銑刀的殘影影像，以次像素計算用來量測微型銑刀的外徑與刃長。至於非魚尾部分微型銑刀瑕疵，本研究擬利用影像重組的方法，重組出一張可用來表示微型銑刀表面的等高線圖，利用影像處理技術來分析此等高線圖，用以檢測出非魚尾部分微型銑刀瑕疵的位置。而魚尾部分的瑕疵，本研究擬採用一聚光光源，放置在一設定的角度上，以凸顯魚尾刃面的瑕疵，以此來檢測魚尾部分的瑕疵。

This project is to develop a machine vision system for the measurement and inspection of the micro-router. A kind of diamond-patterned micro-router which formed by inter-woven spiral blades on the micro-router』s cylindrical surface is the target to be inspected. There are two critical issues to be processed. The first issue is about the speed and accuracy in measuring both width and length of the micro-router. The second issue is about the defects of micro-routers, which will affect the product quality. To solve the first issue, some laser system is traditionally used. However, it is quite expensive and difficult to be adjusted or aligned automatically. For the second issue, human inspection is commonly adopted by the aid of microscope. But it is slow and unstable. So, a machine vision system, including a set of AOI hardware and software algorithm, is proposed and to be developed for auto-measuring and inspecting the micro-router in this project. We divide the proposed system into two sub-systems, including auto-measurement and inspection sub-system. In the auto- measurement sub-system, we attempt to use the minimum gray-value image of contour images to measure the diameter and width of a micro-router in sub-pixel accuracy. In the micro-router inspection sub-system, we attempt to utilize surface reconstruction to reconstruct a set of isograms for representing the surface of a micro-router. And then analyze the isograms by image processing technology to identify the position of the flank-side defect. We attempt to use the spotlighted illumination to highlight the front side defect. Finally, the position of defect will be identified according to the scanning result in polar coordinates. The proposed AOI system can be used as an assistant tool for micro-router quality assurance.