在矽晶太陽能電池製造產業中，成品的數量計算一直是耗費大量人力與時間的一個瓶頸，為了增加產能與降低成本，利用自動化機器來取代原本的人力相當必要的。我們發展了一套自動化視覺數片系統，利用電腦視覺的方式取代原本由人工進行的數量檢測。電腦視覺檢測的內容分為三大部分，第一部份即是透過Contra-harmonic mean filter濾除雜訊與使用改良的直方圖等化來增強影像中表面與縫隙的對比度；第二部份是利用一階導數的特性將可能存在邊緣的特徵區段標示出來；第三部份即是進行邊緣偵測的工作，在標示的區域範圍內搜尋真正為邊緣的地方，並且利用高頻的特性濾除錯誤的偵測線段。最後以偵測到的邊緣個數作為計算片數的結果。

We develop an automatic visual counting system. It replaces conventional manual inspection by automatic computer vision method. The visual inspection composes of three parts: The first one is removing noises via contra-harmonic mean filter and improving the contrast between surfaces and gaps in an image by using specified histogram equalization. The second one is using characteristic of first derivative to mark the region where edges may exist. The third one is about edge detection. We search for real edges in the previously marked region and using characteristic of high frequency to filter out the erroneously detected edges. Finally, we output the counting result of detected edges as the number of solar cells.

[1] Rafael C. Gonzalez, and Richard E. Woods, “Digital Image Processing 2nd”, Prentice-Hall, 2002.
[2] Martin A. Green “Third Generation Photovoltaics : Advanced Solar Energy Conversion”, Springer Berlin Heidelberg, 2006.
[3] Xiaoyi Jiang and Horst Bunke, ”Edge Detection in Range Images Based on Scan Line Approximation”,Computer Vision and Image Understanding Vol. 73, No. 2, February, pp. 183–199, 1999.
[4] Andrea Polesel, Giovanni Ramponi, and V. John Mathews, ”Image Enhancement via Adaptive Unsharp Masking” IEEE Transaction on Image Processing, Vol. 9, No. 3, March 2000.
[5] Jinbo Wu, Zhouping Yin, and Youlun Xiong “The Fast Multilevel Fuzzy Edge Detection of Blurry Images” IEEE Signal Processing Letters, Vol. 14, No. 5, May 2007.
[6] 劉佳怡 全球太陽光電潛力與主要市場, 2007.
[7] 劉佳怡 太陽光電技術發展觀測, 2007.
[8] 劉佳怡 台灣太陽光電產業優勢領域競爭力評估, 2007.