Please use this identifier to cite or link to this item: http://hdl.handle.net/11455/1633
標題: 螺紋自動光學檢測
Automatic Optical Inspection of Screw
作者: 李政翰
Lee, Cheng-Han
關鍵字: computer vision;電腦視覺;three-wire method;Parallel -back light source;binarization;LoG;linear interpolation;三線法;平行光背光源;二值化;LoG;線性內插
出版社: 機械工程學系所
引用: [1]John Fox,“Robust Regression,”January(2002). [2]Qinghang He, Zhenxi Zhang, A new edge detection algorithm for image corrupted by White-Gaussian noise, China,Int. J. Electron. Commun. (AEÜ) 61 (2007) 546 -550. [3]Lijun Ding,Ardeshir Goshtasby,On the Canny edge detector, Dayton, Pattern Recognition 34 (2001) 721-725. [4]Ramesh Jain, Rangachar Kasturi, Brian G.Sshunck, “Machine Vision,”McGraw-Hill International Editions(1995). [5]Zhengyou Zhang,“A Flexible New Technique for Camera Calibration, ” (1998). [6] Dong-Su Kim, Wang-Heon Lee, In-So Kweon, Zhenxi Zhang, Automatic edge detection using 3 • 3 ideal binary pixel patterns and fuzzy-based edge thresholding, South Korea, Pattern Recognition Letters 25 (2004) 101–106. [7] Du-Ming Tsai, A fast thresholding selection procedure for multimodal and unimodal histograms, R.O.C, Pattern Recognition Letters 16 (1995) 653-666. [8]范光照、張郭益,"精密量測",高立圖書,台北市 [9]史詠傑,推拔管螺紋於三次元量床檢測與不確定度分析,國立中興大學機械工程研究所碩士學位論文,台中,2008。 [10]施佑宗,應用機械視覺於檢測內螺紋之可行性探討,國立中 興大學機械工程研究所碩士學位論文,台中,2007。 [11]歐陽衡,機器視覺用在封裝IC外觀瑕疵檢測之研究,國立成功大學工程科學研究所碩士學位論文,台南,2007。
摘要: 
如今的檢驗方式通常採用電腦視覺的方式,利用改良光源、攝影機取像,搭配電腦運算,然後在影像上分析我們的工件,並判斷工件是否合乎標準。
本文採用三線法的原理,透過移動螺紋並記錄其位移量,在螺紋上下半部的相對位置各拍一張影像來做分析,為了避免螺紋表面上的反光或是影像亮度不均,因此採用平行光背光源,接著本文分別對待測物之影像使用二值化與Laplacian of Gaussian(LoG)影像預處理,再透過OPENCV裡的影像處理和線性內插搜尋輪廓座標點並代入線性回歸計算螺紋中每條齒腹所匹配的最佳直線,透過演算法分析其螺紋節距、節徑、大徑與小徑。
最後利用光學投影機重複量測螺紋,與本文實驗結果做比對,再針對實驗誤差因子的影響做不確定分析。

Today inspection method usually adopts computer vision. That is the workpiece is analyzed in the workpiece picture, and it is determined whether the workpiece Conforms with the standard by improving the light source, taking the workpiece picture, and matching up computer''s calculation.
In this paper, we adopt three-wire method that takes down quantity by moving the screw and takes the workpiece picture on lower and higher the screw part .Then we adopt Parallel -back light source for avoiding light reflection or uneven brightness. We use binarization algorithm and LoG algorithm image pretreatment separately and afterward use OPENCV and linear interpolation search edge point which substitutes linear regression to calculate the best matched line of tooth flank which analyze pitch, pitch diameter, big diameter, small diameter in the screw.
Finally, We use profile projector to measure the screw repeatedly that compare with ours experiment''s result in this paper, then that directed agains error factors to do uncertainty analysis.
URI: http://hdl.handle.net/11455/1633
其他識別: U0005-2208201115261700
Appears in Collections:機械工程學系所

Show full item record
 
TAIR Related Article

Google ScholarTM

Check


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.