應用機器視覺技術即時檢測LED之基底瑕疵

Application of Machine Vision Technology on Real-time Detection of LED Defects

劉瑞青(Rui-Jing Liu)

指導教授 ： 田方治

國立臺北科技大學/管理學院/工業工程與管理研究所/碩士(2010年)

https://doi.org/10.6841/NTUT.2010.00269

摘要

超高亮度LED燈被視為第四代光源加上近幾年環保意識提升使的超高亮度LED燈的需求量大增，傳統的人工目視檢測基底以無法滿足產能的需求，本研究主要希望建立一套超高亮度LED基底瑕疵檢測系統以取代原先費時的人工目視檢測以滿足產能需求。本研究研究對象為超高亮度LED燈下游封裝產業所使用的封裝後導熱塊(Slug)亦即基底，主要將基底瑕疵分為六種，透過ROI、簡單的影像處理、Blob分析來達到快速檢測的需求並且將各瑕疵分類。
傳統的最大內切圓求解常會透過CNC機台測量物件座標，並將所得之資料進行最大內切圓求解，本研究提出一個新的最大內切圓(MIC)求解以期能應用在影像處理上，以無理論誤差的方式測量基底製程元件。

關鍵字

瑕疵檢測 ； 快速連通標號 ； 型態學 ； Blob分析 ； K-Curvature ； 最大內切圓 ； Range of interest(ROI)

並列摘要

The brightest LED lamp is regarded as the fourth-generation illuminant. In recent years, due to the raise of environmental awareness, the requirement of the brightest LED lamp has increased. On the contrary, the traditional way of eyeballing the substrate cannot fulfill the requirement of the capacity. Therefore, the main goal of this research is to establish a set of system – the examination of defect in substrate by brightest LED lamp. This device is to replace the original time-consuming eyeballing and to fulfill the requirement of capacity. The object of this research is the packaged slug, which is substrate, used by the downstream packaging industries. The defect in substrate can be separated into six categories. We can categorize these defects by ROI, simple image processing and Blob analysis to fulfill the requirement of rapid examination.
The traditional solution of maximum inscribed circle (MIC) usually uses the CNC to measure the coordinate of object and deal with these measuring data to process the solution of maximum inscribed circle. However, this research brings up a new solution of maximum inscribed circle in expectation of applying to the image processing in the way without the theoretical error to measure the substrate.

並列關鍵字

Defect Detection ； Blob Analysis ； Morphology ； K-Curvature ； Range of interest(ROI) ； Fast Connected Component Labeling ； maximum inscribed circle

參考文獻

[4] Dongik Shin Changsoo Han,a Young Shik Moonb, “Ball stud inspection system using machine vision,”ISA Transactions, Vol.41, 2002, pp. 3-11.

[5] E.S.Gadelmawla , “Simple and Efficient Algorithms for Roundness Evaluation from the Coordinate Measurement Data,” Measurement,Vol.09,2009,pp.3-15.

[6] F. Lisa-Mingo, and J. Carrabina, “On the Application of Configurable Computing to Real-time Industrial Visual Inspection,”Emerging Technologies and Factory Automation, Vol.01,1999,pp.487-494.

[7] G.L.Samual，M.S.Shunmugam，“Evaluation of circularity from coordinate and form data using computational geometric techniques,”Precision Engineering ,Vol.24,2000,pp.251-263.

[8] H. Zheng, L.X. Kong, S. Nahavandi “Automatic inspection of metallic surface defects using genetic algorithms,”Journal of Materials Processing Technology, Vol. 125-126, 2002, pp.427-433.