簡易檢索 / 詳目顯示

研究生: 黃郁倫
Huang, Yu-Lun
論文名稱: 利用彩色正交條紋於條紋反射法之形貌量測
Using Colorful Orthogonal Grid Fringes in Fringe Reflection Method for Specular Surface Profile Measurement
指導教授: 陳元方
Chen, Terry Yuan-Fang
學位類別: 碩士
Master
系所名稱: 工學院 - 機械工程學系
Department of Mechanical Engineering
論文出版年: 2020
畢業學年度: 108
語文別: 中文
論文頁數: 124
中文關鍵詞: 條紋反射法鏡面形貌法向量式彩色正交條紋
外文關鍵詞: Fringe reflection method, Specular profile, Normal vector method, Colorful orthogonal grid fringes
相關次數: 點閱:133下載:0
分享至:
查詢本校圖書館目錄 查詢臺灣博碩士論文知識加值系統 勘誤回報
  • 本文主要利用彩色正交條紋影像的特性,透過簡易的判別式便可將這張條紋影像分解出具三步相位移之x、y方向的二元條紋影像。此方法應用於條紋反射法便可省去大量的拍攝相位移影像時間,相機於一個位置只須拍一張影像就可同時解出x、y方向之相位值,使法向量式系統的拍攝試件過程不再繁雜。
    彩色正交條紋應用於條文反射法中,首先須克服的是實際由螢幕投影出之條紋影像並不像模擬的情形容易,本文透過中值濾波、伸張強化、影像分割法與形態學的一系列影像處理分解出較平整的x、y方向條紋影像,並將解出之相位值實際套用至條紋反射法-法向量系統中,解平面鏡之形貌。
    實驗量測出試件形貌,並拉出其水平曲線與垂直曲線與基準高度值為0的高度值比較。均方根誤差為0.342mm與0.445mm。而改以傅立葉轉換進行二元條紋濾波所解出之相位值帶入法向量式系統量測,其均方根誤差為0.324mm與0.41mm。最後將斜率值抹平後再進行量測,水平與垂直之均方根誤差為0.254mm與0.303mm。於最後探討其量測產生誤差的原因。

    The main purpose of this study is to measure the specular profile using colorful orthogonal grid fringes. Through its characteristics, a simple discriminant can decompose this fringe image into a binary fringe image in the x and y directions with a three-step phase shift. This method is applied to the fringe reflection method can save a lot of time for shooting phase shift images. The process of shooting specimens is no longer complicated.
    First we have to overcome is that the actual fringe image projected by the screen is not as simply as the simulated situation. We use a series of image processing such as median filter, stretching ,image segmentation and morphology to decompose the relatively flat x and y direction fringe images, and actually apply the resolved phase to the fringe reflection system.
    The specular surface profile was measured, and the horizontal curve and vertical curve were drawn to compare with the height value of the reference height value of 0. The root-mean-square error (Vertical direction) is found to be within 0.342 mm. The horizontal direction is found to be within 0.445 mm.
    Then we change the method to FFT filter to solve the phase noise.The root-mean-square error (Vertical direction) is found to be within 0.324 mm. The horizontal direction is found to be within 0.41 mm.
    Last we use mean filtering to smooth the noise of the slope images. The root-mean-square error (Vertical direction) is found to be within 0.254 mm. The horizontal direction is found to be within 0.303 mm.
    Keywords:fringe reflection method, specular profile,normal vector method, colorful orthogonal grid fringes

    摘要 I 目錄 VII 表目錄 X 圖目錄 XI 第一章 緒論 1 1.1 研究背景 1 1.2 研究目的 3 1.3 文獻回顧 4 1.4 本文架構 6 第二章 條紋反射法之原理 7 2.1 條紋反射法-法向量式系統 7 2.1.1 攝影機鏡心位置測定方法[1][3][4] 9 2.1.2 反射向量表示式 20 2.1.3液晶螢幕之定位方法[1] 21 2.1.4 系統中的法向量計算方法[1] 31 2.1.5 法向量轉換斜率積分方法 35 2.2 彩色正交條紋解相位原理[13] 37 2.2.1 正交條紋解相位原理 37 2.2.2 彩色正交條紋影像 39 第三章 影像處理與彩色正交條紋相位解析法 40 3.1 彩色正交條紋相位解析 40 3.1.1伸張強化 41 3.1.2 中值濾波(Median Filter)[8] 43 3.1.3 形態學[14] 44 3.1.4 臨界值法[14] 48 3.1.5 三步相位移法[16] 51 3.1.6 Macy相位展開法[15] 54 3.2 影像平面之幾何校正方法 55 3.2.1 應用形狀函數校正影像[5][6][7] 55 3.2.2 影像空格灰階填補[5] 64 第四章 實驗架設與系統校正 66 4.1 法向量式系統校正 66 4.2 法向量式實驗架設 75 第五章 實驗結果與討論 83 5.1 彩色正交條紋模擬 83 5.1.1 將四元正交條紋組合出彩色正交條紋 83 5.1.2 由螢幕投影出之彩色正交條紋 89 5.2 法向量式量測 99 5.2.1 平面鏡量測 99 5.3 以傅立葉轉換解相位展開 110 5.4 斜率抹平後量測 116 第六章 結論與建議 119 6.1 結論 119 6.2 建議 121 參考文獻 122

    1.蘇省任,“具大曲率之反射性表面形貌量測”,國立成功大學機械工程學系碩士論文,2013。
    2.黃博偉,“應用條紋反射法量測鏡面形貌”,國立成功大學機械工程學系碩士論文,2019。
    3.R. W. Webster and Y. Wei, ““ARINE P.” – A Robot Golfing System Using Binocular Stereo Vision and a Heuristic Feedback Mechanism,” Proceeding of the 1992 IEEE/RSJ International Conference on Intelligent Robots and System, pp. 2027-2034, 1992.
    4.黃稚桓,“具雙重模式光學三維量測系統之建構”,國立成功大學機械工程學系碩士論文,2009。
    5.陳延昌,“應用幾何校正於條紋反射法作物體外貌量測”,國立成功大學機械工程學系碩士論文,2008。
    6.Yuan Y. Tang and Ching Y. Suen, “Image Transformation Approach to Nonlinear Shape Restoration,” IEEE Systems, Man, and Cybernetics Society , Vol. 23, No.1, 1993.
    7.Singiresu S. Rao, “The Finite Element Method in Engineering,” Oxford, New York, 2004.
    8.繆紹綱,“數位影像處理”,台灣培生教育出版股份有限公司,2003年。
    9.William W. Macy, “Two-dimensional fringe-pattern analysis,” Applied Optics, Vol. 22, No.23, pp. 3898-3901, 1983.
    10.Markus C. Knauer, Jürgen Kaminski, and Gerd Häusler, “Phase Measuring Deflectometry: a new approach to measure specular free-form surfaces,” Proceeding of SPIE Vol.5457, pp.366-376, 2004.
    11.Tian Zhou, Kun Chen, Haoyun Wei, and Yan Li, “Improved method for rapid shape recovery of large specular surfaces based on phase measuring deflectometry,” Applied Optics, Vol. 55, No. 10, 2016.
    12.Yan Tang, Xianyu Su, Yuankun Liu, and Hailong Jing, “3D shape measurement of the aspheric mirror by advanced phase measuring deflectometry,” OPTICS EXPRESS, Vol. 16, No.19, 2008.
    13.Xu, Xueyang, Zhang, Xiangchao, Xu, Min,Dynamic “three-dimensional shape measurement for specular freeform surfaces with the quaternary orthogonal grid fringes”,SPIE 2016.
    14.黃俊銘,“應用數位影像處理於焦炭粒度檢測”,國立成功大學機械工程學系碩士論文,2009。
    15.Macy, W.W., “Two-dimensional fringe-pattern analysis,”Applied Optics, 1983. 22(23): p. 3898-3901.
    16.徐弘峻,“窗型傅立葉濾波與隨機相移法於干涉顯微鏡應用” 國立中興大學機械工程學系碩士論文,2014。
    17.Hongwei Zhang, Shujian Han, Shugui Liu, Shaohui Li, Lishuan Ji, and Xiaojie Zhang, “3D shape reconstruction of large specular surface”, Processing of SPIE ,2012.

    18.ZHONGQUAN Wu,LINGXIAO Lx, “A Line-Integration Based Method for Depth Recoveryfrom Surface Normals”,1988.
    19.Xie, Pei, Tang, Minxue, Wei, Xiaoru, “Three-dimensional shape measurement of specular surfaces by orthogonal composite fringe reflection”, Processing of SPIE ,2011.
    20.Shuangyan Lei,Song Zhang, “Flexible 3-D shape measurement using projector defocusing”,OPTICS LETTERS,2009.
    21.Laura Ekstr,Song Zhang, “Three-dimensional profilometry with nearly focused,binary phase-shifting algorithms”, OPTICS LETTERS,2011.
    22.B¬O LI,SUODONG MA,YANG ZHAI, “Fast temporal phase unwrapping method for the fringe reflection technique based on the orthogonal grid fringes”,Applied Optics,2015.

    下載圖示 校內:2025-09-03公開
    校外:2025-09-03公開
    QR CODE