201237359 六、發明說明： 【發明所屬之技術領域】 本發明係有關於一種光學式強度型三維表面形貌與顯微量測裝置 及方法，尤指一種利用角度偏向與光學成像技術，藉由光感測陣列元件 取出影像，由其二維圖像測量其面積尺寸大小，並由其光強度變化(反 射率變量）轉換成高度值，而成三維之圖像之技術。 【先前技術】 按目前所知表面姆度之測量，由於早期表面精密度要求不是那 麼嚴謹’大娜取目視方法。但近來科技進步快，對於表面精密度的 要求也越來越高了，為了加快線上檢驗速度錢化儀^的複雜度光 強度法是比較好的選擇’由於架構與原理簡單，不管是光學系統的架 5又、調整與；k正上’都很容易…般而言’在量測上可分為接觸式跟 非接觸式二大類，接觸式的量測如傳統的接觸探針式㈣心沉如 instrument)。為了得到較準麵測量值，一般是在制物上之不同位 置作了幾次_量後，再計算出平均值，不過這需要㈣相當多的時 間，且對軟質材料而言，恐有到傷其表面之虞。所以本發明提出非接 觸式的制方法可以在赠_作麵_作外，也不會破壞到待測 物上的表面，影響到準確度。 由於越來越多人提出三維表面的量測方法所以證年㈤卿 等人[如參考文獻6]提出—種數微鏡裝置（digital —device)技術的顯微像位移輪廓儀，該數字微鏡裝置是連 同其照明光學組合成-種三維立體顯微鏡，主要由計算產生的正弦光 強分佈之條紋圖案通過顯纖，然後投影到要量測的物體表面上。由 於物體表面的雜使得純產以變之後關由電軸 3 201237359 (Charge-Coupled Device，CCD)記錄起來，之後再利用相位移的技術 重建物體表面的顯微三維形貌。其實驗結果成功證明具備這種測量粗 糙表面的技術之表面輪廓在微米等級。 2001年Cz. Lukianowicz等人[如參考文獻7]提出非接觸之反射式 的量測法量測粗糙度，且量測時以移動的方式，或旋轉的方法作量測， 之後再由CCD擷取相對的光強度對像素的變化，即可得到粗輪度或形 貌，不過此方法只以一維的方法呈現。 近期的2010年*在工業用途中’基於快速、準確地測量表面形貌、 距離和厚度之原則下，Antonin Miks等人[如參考文獻1]所示提出非 接觸光學測量像共焦理論，這些感測器是基於光學系統縱向色差波長 的相依性，主要疋將測罝樣品的距離或厚度編碼成光譜資訊。此方法 是非接觸式測量平行板的厚度，由於透明的樣品會有材料色散和畸變 等問題，此量測誤差是以量測CCD光強度的變化，再作數值分析之後加 以修正，得到待測物厚度跟距離。 同年YajimWang等人[如參考文獻2]提出無相位展開的方式，以調 整後的傅立葉(Fourier)轉換法量得三維的表面輪廓，主要架構是以投 影機投射在待測物上之後再藉由CCD擷取未被展開相位之光強度變 化，再結合由顏色資訊計算出相位移範圍，即可取得真正的相位值。 2006年，Oleg V. Angelsky等人[如參考文獻3]也是利用光學上非 接觸的技術來測量粗縫的不均勻表面，主要架構是利用㈤耐·，s interferometer)產生光程差’織使肌⑶拍攝干涉賴形，再重建 大面積的表面輪廓圖。 2005年，Madhuri Thakur等人[如參考文獻4]利用化触 mterfietry量測透明待測物。主要架構是利用準直雷射光穿透一 [S ] 4 201237359 個光柵，產生具有週期性的圖形，之後再穿透待測物，最後再經過一 光拇互相干涉即產生托伯干涉條紋（Talbot interferometric fringes)。然後利用CCD拍攝干涉條紋，最後實驗結果同樣以探針的量 測結果做比較。 2005年’ Jeong Seok Oh等人[如參考文獻5]提出飛秒雷射脈衝 (Femtosecond laser pulses)表面輪廓量測儀，且討論了兩種方法來 使用飛秒脈衝雷射當光源增強干涉，增強精密表面剖面測量。第一種 是以一超抉激光脈衝重複低，時間連續掃描干涉，原本這干涉術不適 用於白光’但經改良後可使自絲這干涉術可有效細；第二為飛秒 脈衝雷射的高空間關性’以—翻對小型的參考面，可以更大的光 學測試在非對稱配置。這兩個優勢由Fizeau及Twyman _ Green型掃描干 涉實驗驗證。 綜整目前所知之表面形貌量測技術，尚無一種可快速且穩定及低 成本量測裝置可嘉惠產業。為了幫助精密的工業量測能賊有效快速 的得到想要的大範圍表面形貌以及降低成本，鑑於雷射在光學的量測 上速度快以及穩定等優點，且目前有關這方面技術之論文及專利並不 多見。本發明人乃積極投入研究，在國科會計劃補助下，經不斷地實 驗與研發’終有本發明成果產出。為求研發成果及心血獲得合法的權 益保障，乃依法具文提出專利申請。 【發明内容】 本發明之主要目的，在於提供一種光學式強度型三維表面形貌與 顯微量測裝置，其為一種以顯微的方法結合臨界角強度法和CCD光強 度影像擷取來做為缺陷量測裝置。此技術可以讓精密量測產業與其他 相關產業的發展速度快速提升。在操作上，因為可以簡化很多操作上 [S ] 5 201237359 的不方便，所以在學習操作上可以很快速的熟悉，而且因本發明為非 接觸式的量測，在量測時不會造成量測上的損失，並避免不必要的時 間跟金錢的浪費。 達成前述目的之技術手段，係提供有光源、擴束器、偏極板、第 二透鏡、偏極分光鏡、第二物鏡、旋轉平台、四分之一波片、第二透 鏡、角度感測器及矩陣式光感測器。以光源發射一光束；光束經擴束 器擴大為平行光束；平行光束經偏極板成為P偏振光；P偏振光光束 穿透第二透鏡與偏極分光鏡後緊接著經過該四分之一波片成為圓偏 極’並經第二物鏡而投射至一待測物上；第二物鏡供來自待測物之反 射光穿越返回偏極分光鏡，以放大待測物表面形貌；以第三透鏡供自 偏極分光鏡返回的反射光通過；以角度感測器在臨界角附近做為微小 角度感測；及以矩陣式光感測H拍攝及擷取待測物反射光之光強度變 化’並傳送至一電腦進行處理，由所擷取之二維影像測量其面積尺寸 大小，並由其強度變化(或反射率變量）而轉換成高度值，進而形成三維 之影像。 【實施方式】 壹•本發明之技術構想 本發明之主要構想，侧用光學快速量測之伽，錢架構簡單 化便且化操作單純化等作為考量。本發明係將光強度量測的方法 結合臨界角與CCD擁取影像技術，即時利用待測物反射回CCD的光強度 變化’迅速轉換成3D形貌’從3D圖我們可以即時的得知我們想要的待 測物表面的缺陷以及粗較度變化情形。而且，本發明架構也具有顯微的 效果’在制上也具有高錄度、高解析度等優點。 由於需要在全反射與臨界角附近來回轉動稜鏡且希望所得到， 6 201237359 兩個圖缝完全重疊’才能真正算出此待咖的反神分布圖，機構 上的誤差可能會因為長時間的運動而造成量測上的誤差，因此也必須 有自動校正的功能。 貳·本發明之技術原理 1·平行四邊形稜鏡入射外角與反射率之關係 本發明之架構’主要是以s偏振光作為平行四邊形稜鏡内部的反射 率變化’由於S偏振光在平行四邊形稜鏡内部經過兩次的反射之後可以 得到較為接近線性的曲線，之後再取其斜率之平均值，由於斜率與平 行四邊形稜鏡的反射率以及平行四邊形稜鏡的外部入射角0之間有相 對的關係’本發明人乃將之加以應用。本發明人首先研究如何求得反 射率’以及反射率與平行四邊形稜鏡的外部入射角0之間的關係式。根 據成像系統，主要是利用ABCD矩陣來作為追求光軌跡的變化外，如 離軸高度以及光線賴斜肖度’紐再縣發顺求得之反射率對表 面尚度的變化作進一步結合，即可利用CCD所攝取到的圖片來作為得 到三維表面高度變化的輝。所以在祕系財，本發明會推導待測 物表面的高度變倾平行四邊賴鏡人射肢的_，飾求得平行 四邊形稜鏡反射率與待測物表面的關係式。由此得知，只要知道反射 率的變化，即可知道待測物的表面高度。在量得反射率的同時必需考 慮到光強度的變動，所以對於CCD晝面的攝取要很謹慎，因此重複性 的晝面攝取是必要的’最後再取其光強度平均值來作為求取表面高度 的依據。本發明之技術有如顯微鏡的功能，不但可以看清楚待測物的 表面輪廓之外，還可以求得三維的表面變化，所以本發明架構不但有 其量測3D之效果，亦可做為表面檢驗，如粗糙度、表面形貌或缺陷等 量測。 201237359 如圖1所示，反射光於平行四邊形棱鏡200内部作兩次之反射。由 邊界條件以及Snell’s法則(law)可求出兩次反射之反射率對外角^之 關係式’如式(1)及式(1-1)與式(1_2)所示： «1 COS0 -«2 cos ^2 (1)201237359 VI. Description of the Invention: [Technical Field of the Invention] The present invention relates to an optical intensity type three-dimensional surface topography and microscopic measuring device and method, and more particularly to an optical focusing technique and optical imaging technology, by means of light The technique of sensing an array element to take out an image, measuring its area size by its two-dimensional image, and converting it into a height value by changing its light intensity (reflectance variable) into a three-dimensional image. [Prior Art] According to the current measurement of the surface roughness, the early surface precision requirements are not so rigorous. However, recent advances in science and technology have made the requirements for surface precision higher and higher. In order to speed up the speed of online inspection, the complexity of the light intensity method is a good choice. 'Since the architecture and principle are simple, no matter the optical system. Rack 5, adjustment and; k is on the 'very easy... Generally speaking, the measurement can be divided into two types of contact and non-contact, contact measurement such as the traditional contact probe type (four) heart Shen as instrument). In order to obtain the quasi-surface measurement value, the average value is usually calculated after several times at different positions on the workpiece, but this requires (four) a considerable amount of time, and for soft materials, it is feared that Injury to the surface. Therefore, the present invention proposes that the non-contact method can be applied to the surface of the object to be tested without affecting the surface. As more and more people propose three-dimensional surface measurement methods, Zhengnian (5) Qing et al. [such as Reference 6] proposed a micro-image displacement profiler with digital micro-device technology. The mirror device is combined with its illumination optics to form a three-dimensional stereo microscope. The stripe pattern of the sinusoidal light intensity distribution generated by calculation is passed through the fiber and then projected onto the surface of the object to be measured. Due to the impurity on the surface of the object, the pure product is changed and then recorded by the electric axis 3 201237359 (Charge-Coupled Device, CCD), and then the phase displacement technique is used to reconstruct the microscopic three-dimensional shape of the surface of the object. The results of the experiment have successfully demonstrated that the surface profile of this technique for measuring rough surfaces is on the order of microns. In 2001, Cz. Lukianowicz et al. [Reference 7] proposed a non-contact reflective measurement method to measure the roughness, and measure it by moving or rotating, and then by CCD. A coarse rotation or topography can be obtained by taking the relative light intensity versus the pixel, but this method is only presented in a one-dimensional manner. In the recent 2010* in industrial applications, based on the rapid and accurate measurement of surface topography, distance and thickness, Antonin Miks et al. [Reference 1] proposed non-contact optical measurement like confocal theory. The sensor is based on the dependence of the longitudinal chromatic aberration wavelength of the optical system, and mainly encodes the distance or thickness of the measured sample into spectral information. This method is a non-contact measurement of the thickness of the parallel plate. Since the transparent sample may have problems such as material dispersion and distortion, the measurement error is to measure the change of the CCD light intensity, and then perform numerical analysis to correct the object to be tested. Thickness and distance. In the same year, Yajim Wang et al. [J. 2] proposed a phase-free expansion method to measure the three-dimensional surface profile by the adjusted Fourier transform method. The main structure is projected by the projector on the object to be tested. The CCD captures the intensity change of the unexpanded phase, and combines the color information to calculate the phase shift range to obtain the true phase value. In 2006, Oleg V. Angelsky et al. [see Reference 3] also used optical non-contact techniques to measure the uneven surface of the coarse seam. The main framework is to use (5) s interferometer to produce optical path difference The muscle (3) captures the interference and reconstructs the large-area surface contour map. In 2005, Madhuri Thakur et al. [Reference 4] used a chemo-mterfietry to measure transparent analytes. The main structure is to use collimated laser light to penetrate a [S ] 4 201237359 gratings to produce a periodic pattern, then penetrate the object to be tested, and finally through a mutual interference of the light to produce a Tober interference fringe (Talbot Interferometric fringes). The interference fringes were then taken using a CCD, and the final experimental results were also compared with the measured results of the probes. In 2005, Jeong Seok Oh et al. [Reference 5] proposed a Femtosecond laser pulses surface profile measuring instrument, and discussed two methods to use femtosecond pulsed lasers to enhance interference and enhance the light source. Precision surface profile measurement. The first type is a super-short laser pulse repeating low, time-continuous scanning interference. Originally, this interferometry is not suitable for white light's. However, the improved self-wire interference can be effectively fined; the second is a femtosecond pulse laser. The high spatiality of the 'to-turn on the small reference plane allows for greater optical testing in an asymmetric configuration. These two advantages are verified by the Fizeau and Twyman _ Green type scanning intervention experiments. In view of the current surface topography measurement technology, there is no one that can be quickly and stably and low-cost measurement devices can be used in the benefit industry. In order to help the sophisticated industrial measurement thief to quickly and efficiently obtain the desired wide-area surface topography and reduce the cost, in view of the advantages of laser speed and stability in optical measurement, and the current papers on this technology and Patents are rare. The present inventors actively participated in the research, and under the program subsidy of the National Science Council, through continuous experimentation and research and development, the results of the invention are finally produced. In order to obtain the legitimate rights and interests of research and development results and efforts, it is legal to file a patent application. SUMMARY OF THE INVENTION The main object of the present invention is to provide an optical intensity type three-dimensional surface topography and microscopic measuring device, which is a microscopic method combined with a critical angle intensity method and a CCD light intensity image capture. It is a defect measuring device. This technology can rapidly increase the speed of the precision measurement industry and other related industries. In operation, since the inconvenience of [S] 5 201237359 can be simplified in many operations, it can be quickly familiarized with the learning operation, and since the present invention is a non-contact measurement, it does not cause an amount in the measurement. Measure the loss and avoid unnecessary time and waste of money. The technical means for achieving the foregoing object is to provide a light source, a beam expander, a polarizing plate, a second lens, a polarizing beam splitter, a second objective lens, a rotating platform, a quarter wave plate, a second lens, and an angle sensing. And matrix light sensors. A light beam is emitted by the light source; the light beam is expanded into a parallel beam by the beam expander; the parallel light beam is P-polarized light through the polarizing plate; the P-polarized light beam passes through the second lens and the polarizing beam splitter and then passes through the quarter The wave plate becomes a circular polarization pole and is projected onto a sample to be tested by the second objective lens; the second objective lens supplies the reflected light from the object to be tested through the returning polarization beam splitter to enlarge the surface topography of the object to be tested; The three lenses are used for the reflected light returning from the polarizing beam splitter; the angle sensor is used as a small angle sensing near the critical angle; and the matrix light sensing H is used to capture and extract the light intensity of the reflected light of the object to be tested. The change is transmitted to a computer for processing, and the size of the area is measured by the captured two-dimensional image, and converted into a height value by its intensity change (or reflectance variable) to form a three-dimensional image. [Embodiment] The present invention is based on the main idea of the present invention, in which the side is optically fast-measured, the money structure is simplified, and the operation is simplistic. The invention combines the method of measuring the light intensity with the critical angle and the CCD image capturing technology, and instantly converts the light intensity change reflected back to the CCD by the object to be tested and quickly converts into a 3D shape. From the 3D image, we can immediately know us. The desired defect on the surface of the object to be tested and the change in the roughness. Moreover, the structure of the present invention also has a microscopic effect, which also has the advantages of high recording, high resolution, and the like. Since it is necessary to rotate back and forth around the total reflection and the critical angle, and hope to get it, 6 201237359 The two seams completely overlap 'can really calculate the anti-divine map of the coffee, the error of the mechanism may be due to the long movement This causes errors in the measurement, so it must also have an automatic correction function.贰·Technical Principles of the Invention 1. The relationship between the incident external angle and the reflectivity of the parallelogram 本 The structure of the present invention 'is mainly the s-polarized light as the parallelogram 稜鏡 inside the reflectance change' due to the S-polarized light in the parallelogram rib After two reflections inside the mirror, a nearly linear curve can be obtained, and then the average value of the slope is taken. Since the slope is opposite to the reflectivity of the parallelogram 以及 and the external incident angle 0 of the parallelogram 稜鏡The relationship 'is the inventor's application. The inventors first studied how to obtain the relationship between the reflectance' and the reflectance and the external incident angle 0 of the parallelogram 稜鏡. According to the imaging system, the ABCD matrix is mainly used as the pursuit of the change of the light trajectory, such as the off-axis height and the ray slanting degree, and the reflectivity obtained by the New York County shun is further combined with the change of the surface scent, ie The picture taken by the CCD can be used as a radiance for obtaining a three-dimensional surface height change. Therefore, in the secret system, the invention will infer that the height of the surface of the object to be tested is tilted parallel to the four sides of the mirror, and the relationship between the reflectivity of the parallelogram and the surface of the object to be tested is obtained. From this, it is known that as long as the change in reflectance is known, the surface height of the object to be tested can be known. In order to measure the reflectance, it is necessary to take into account the variation of the light intensity. Therefore, it is necessary to take care of the CCD surface. Therefore, it is necessary to repeat the facet intake. Finally, the average value of the light intensity is taken as the surface. The basis of height. The technology of the present invention has the function of a microscope, not only can clearly see the surface contour of the object to be tested, but also can obtain a three-dimensional surface change, so the structure of the invention not only has the effect of measuring 3D, but also can be used as a surface inspection. Such as roughness, surface topography or defects. 201237359 As shown in FIG. 1, the reflected light is reflected twice inside the parallelogram prism 200. From the boundary conditions and the Snell's law, the relationship between the reflectance of the two reflections and the external angle ^ can be obtained as shown in equations (1) and (1-1) and (1_2): «1 COS0 -« 2 cos ^2 (1)
«j COS^j + «2 COS 45° - sin' sin^> sin θ】 Ο 1 «Λ 0X11 η. (1-1) (1-2) 由式(1)，我們可以利用MATLAB軟體（MATLAB是MATrixLABoratory 的縮寫，是由美國MathWorks公司出品的商業數學軟體）模擬出如圖2 所示之結果。不過為了更能證明此曲線之方程式的正確性，由圖3之實 際量測平行四邊形之反射率曲線，加以比較之。因此，由圖3我們可以 了解到最大的斜率約在5.608度，此角度為臨界角ec(此時角度是指外角 Θ )’該附近角度之反射率變化是最靈敏的。由於本發明之架構是以s 偏極作為主要平行四邊形棱鏡之反射率，所以量測上可以比p偏極有較 大的高度量測範圍，因為線性的範圍較大，可以有較大的入射外角㊀。 2.成像系統與反射率之關係 本發明之架構中的成像系統如圖4所示，主要是由一個第二物鏡 (0bjective2)17與一個第三透鏡(Lens3)i8所組成，最後成像於電荷耦合 器CCD (Charge-Coupled Device)上。因此我們可以根據abCD矩陣，如 式(2)所示，最後我們可以得到如式(3)所示之角度4χ(變化關係與 式(4)所示之離光軸距離λ:3、夂之關係： 201237359 A. d2 let's άχ =/!+/2 /2 0 1 dx ΙΑ 1 xs f? foi/c τ ·ίγ+/ι+/2'|^2 f2 ifds=fx = /2«j COS^j + «2 COS 45° - sin' sin^> sin θ】 Ο 1 «Λ 0X11 η. (1-1) (1-2) From equation (1), we can use MATLAB software ( MATLAB is the abbreviation of MATRIxLABoratory, which is a commercial mathematics software produced by MathWorks, USA. It simulates the results shown in Figure 2. However, in order to better prove the correctness of the equation of this curve, the reflectance curve of the parallelogram is actually measured by Fig. 3 and compared. Therefore, we can see from Fig. 3 that the maximum slope is about 5.608 degrees, which is the critical angle ec (the angle is the outer angle Θ). The reflectance change of the nearby angle is the most sensitive. Since the architecture of the present invention uses the s-polarity as the reflectivity of the main parallelogram prism, the measurement can have a larger height measurement range than the p-polarity, because the linear range is larger, and the incident can be larger. One outer corner. 2. Relationship between imaging system and reflectivity The imaging system in the architecture of the present invention is shown in Fig. 4, mainly composed of a second objective lens (17) and a third lens (Lens3) i8, and finally formed on the charge. Coupler CCD (Charge-Coupled Device). Therefore, we can according to the abCD matrix, as shown in equation (2), and finally we can get the angle 4χ as shown in equation (3) (the relationship between the change and the distance from the optical axis shown in equation (4) is λ: 3, 夂Relationship: 201237359 A. d2 let's άχ =/!+/2 /2 0 1 dx ΙΑ 1 xs f? foi/c τ ·ίγ+/ι+/2'|^2 f2 ifds=fx = /2
X (2)X (2)
⑶ (4) 從 '()中T 乂决疋冨入射於平行四邊形稜鏡pHsm) 時之入射肖度，不俩核射之_考慮義極分光鏡(pGlarizing Beamsplitter，)所造成之角度變化’所以式⑶還需乘於—個負號， 也就疋令气-A，A為真正入射於平行四邊形稜鏡之角度變化量。 從式(4)帽可以決定我們想的放大倍率。圖5所示，為表面高度之幾 何圖形。我們將由近似的方法求得式(5)，且代入式⑶得到式⑹角度對 高度之變化。 (5) ⑹ χ' = 2ΑΘ « _2-— s Δχ(3) (4) The angle of incidence from the '() T 乂 疋冨 疋冨 疋冨 疋冨 疋冨 疋冨 疋冨 疋冨 稜鏡 稜鏡 稜鏡 , , , , , , , , , , 考虑 考虑 考虑 考虑 考虑 考虑 考虑 考虑 考虑 考虑 考虑 考虑 考虑 考虑 考虑 考虑 考虑 考虑 考虑 考虑 考虑 考虑 考虑 考虑 考虑 考虑 考虑 考虑 考虑 考虑 考虑 考虑Therefore, equation (3) also needs to be multiplied by a negative sign, which means that the gas-A, A is the angular change of the true incident on the parallelogram. From the formula (4) cap can determine the magnification we want. Figure 5 shows the geometry of the surface height. We will obtain equation (5) from the approximate method and substitute equation (3) to obtain the change in angle to height of equation (6). (5) (6) χ' = 2ΑΘ « _2-- s Δχ
Ah =λ：3(·^-Δχ) , ^3<〇 式(7)為平行四邊形棱鏡之入射角度對反射率之斜率關係，因此由 201237359 ()代<(6)我們可以得到式⑼，再由式⑼)簡化成式⑴)所示之反 射率對表面高度變化之義式： m 4>〇,ARs2<〇 n AR.Χ3=~^Γ，m<〇，《=o Ah = ~s2- ⑺ 一 mAh = λ: 3 (·^-Δχ) , ^3 < 〇 (7) is the slope relationship between the incident angle of the parallelogram prism and the reflectivity, so we can get the formula (9) by 201237359 () generation < (6) Then, by the formula (9)), the formula for the change of the reflectance to the surface height shown in the formula (1)) is simplified: m 4 > 〇, ARs2 < 〇 n AR. Χ 3 = ~ ^ Γ, m < 〇, "= o Ah = ~s2- (7) one m
Ah<〇 ⑻⑼Ah<〇 (8)(9)
Ze/、从丄 △/2 = A/^2 · Λ/ M>〇 3·系統架構： (10) (11) 由前面介紹後發現’我們可以利用簡單的架構跟幾何原理來量測 出我們想要的表面高度變化量。域將簡短地介紹_下本發明所使用 的實驗架構’如圖6师。在軟财面主要是制Matiab軟體來計算 出待測物表面高度變化以及祕度，最後顯示出三維_。三維圖形 的好處在於我們可以快速的看出待測物的表面變化情況或缺陷，而且 本架構可以即時的作大範圍的量測，避免時間的等待。 3· 1光學元件及儀器介紹 如圖6所示，本發明一種具體實施例之系統架構，係包括： 光源ίο，其用以發射出光束，其可為單一波長（波長為632 8nm) 之系統光源，或雷射光源，本發明具體實施例係為雷射光源； 光阻隔器（Isolator)ll，其供來自光源的光束經過，用以避免系 統的反射光返回光源，而造成光源的損傷； 201237359 擴束器（Beam Expander)12，其接收來自光阻隔器11光束，用以 將光束擴大為平行光束，本發明一種具體實施例，擴束器12包括一第 一物鏡120、一第一透鏡121，及介於第一物鏡120與第一透鏡121之 間的銷孔122 ;其中第一物鏡120用以使光束經銷孔122後為擴大光 束，其焦距上設一空間濾波器（Spatial Filter)(圖中未示），以便濾 除雜散光源；而第一透鏡121用以使擴大光束成平行光束； 偏極板(Polarizer)13，其具調整角度功能，用以改變來自擴束器 12之光束的光偏極的方向； 第二透鏡(Lens)14，其用以將來自偏極板13之光束擴大及聚焦光 點； 偏極分光鏡(Polarizing Beamsplitter，PBS)15，其用以供來自 該第二透鏡14之光束通過並投射至一待測物30上； 第二物鏡(〇bjective)17，其介於偏極分光鏡15與待測物3〇之 間，其供來自偏極分光鏡15之光束穿越而投射在待測物3〇表面，並 供來自待測物30之反射光穿越返回偏極分光鏡丨5，以作為放大待測 物30表面形貌； 四分之一波片（| Wave plate)16，其介於偏極分光鏡15與第二 物鏡17之間，來自偏極分光鏡15之光束經過該波片16，會變成圓偏 極光； 第二透鏡18，用以供自該偏極分光鏡15返回的該反射光通過； 旋轉平台（Rotation Stage)19，其為一可改變角度的平台； 角度感測器20，其置設在旋轉平台19上，利用旋轉平台19旋 201237359 角度感測器20，使來自第三透鏡18的反射光入射角度感測器2〇做多 次反射’並控制在臨界角附近對反射光做為微小角度感測，本發明一 種具體實施例中，該角度感測器20為一有鍍膜的平行四邊形稜鏡 (Parallelogram Prism) ’其反射次數為兩次；及 矩陣式光感測器21，用來拍攝及擷取待測物反射光之光強度變 化，並傳送至一電腦22進行處理而計算出待測物3〇的表面形貌；本 發明一種具體實施例，矩陣式光感測器21為電荷粞合器 (Charge-Coupled Device ， CCD)。 3.2量測結果 本發明之實驗架構如圖6所示，以氦氖雷射(He-Ne Laser)當作光 源10,經過一光阻隔器（IS0lat0r) u以避免光學系統的反射光返回 雷射内部，再由擴束器（Beam Expander)12擴大為平行光束，之後經 過偏極板14(P (0° ))成為p偏振光，當光束穿透第二透鏡14與偏極 分光鏡(PBS)15後緊接著經過四分之-波長波# 16成為gj偏極，之後 要測量的待測物30樣本(本發明以光栅及又/2〇標準片為主）經光束照 射後’將再經由第二物鏡17、偏極分光鏡15、第三透鏡μ、角度感 測器20成像在電荷搞合器⑽)21上。矩陣式光感測器18(即電荷耦 合器18(CCD))是藉由通用型串列匯流連接璋（USB)直接連至電腦四， 最後可由電腦22上的軟體秀出CCD所拍攝的影像，主要是各別於全反 射(0。)以及臨界肖（5· _。)附近時各拍攝—張灰階之光·變化圖，分 別為I。2以及I。，，然後再利用Matlab軟體作分析，其所拍攝之光強度 變化圖以私2= L/ I。2下去求出二次反射過後之反射率再利用式（^) i S ] 12 201237359 得到高度變化，最後畫出三維輪_。,然而角度感測器2G(即平行四 邊形稜鏡(Parallelogram Prism))是搭配旋轉平台(R〇tati〇n沿喂⑽，利 用旋轉平台19控制角度感測器20來回轉動於全反射角與臨界角之 間。首先以乂/2〇之標準做為系統誤差量挪，由本發明之實驗架構量出 的表面粗糙度值減去由原子力顯微鏡量出的粗糙度值(當作基準值）， 來估计本發明之實驗架構存在的系統誤差之後再利用式(12)、式(^) 修正系統誤差。 3· 2.1非穿透型待測物之量測結果(λ/2〇標準片） 首先將由矩陣式光感測器21為電荷耦合器(Charge_c〇upledZe/, from 丄△/2 = A/^2 · Λ/ M>〇3·System architecture: (10) (11) From the introduction, we found that 'we can use simple architecture and geometric principle to measure us The amount of surface height desired. The domain will briefly introduce the experimental architecture used in the present invention as shown in Figure 6. In the soft finance, the Matiab software is mainly used to calculate the surface height change and the secretness of the object to be tested, and finally the three-dimensional image is displayed. The advantage of 3D graphics is that we can quickly see the surface changes or defects of the object under test, and the architecture can be used for a wide range of measurements in real time, avoiding the waiting time. 3.1 Optical Components and Apparatus Introduction As shown in FIG. 6, a system architecture of a specific embodiment of the present invention includes: a light source ίο for emitting a light beam, which can be a single wavelength (wavelength of 632 8 nm) system. A light source, or a laser light source, is a laser light source; a light blocker (Isolator) 11 for passing a light beam from a light source to prevent the reflected light of the system from returning to the light source, thereby causing damage to the light source; 201237359 Beam Expander 12, which receives a light beam from the light blocker 11 for expanding the light beam into a parallel light beam. In one embodiment of the present invention, the beam expander 12 includes a first objective lens 120 and a first lens. 121, and a pin hole 122 between the first objective lens 120 and the first lens 121; wherein the first objective lens 120 is used to enlarge the light beam after passing the light beam through the pin hole 122, and a spatial filter is disposed on the focal length thereof. (not shown) to filter out the stray light source; the first lens 121 is used to make the enlarged beam into a parallel beam; the Polarizer 13 has an angle adjustment function for changing the beam expander 12 It The direction of the beam's light polarization; a second lens (Lens) 14 for expanding and focusing the beam from the polarizing plate 13; a Polarizing Beamsplitter (PBS) 15 for The light beam of the second lens 14 passes through and is projected onto an object to be tested 30; a second objective lens 17 is interposed between the polarizing beam splitter 15 and the object to be tested 3 The beam of the mirror 15 passes through and is projected on the surface of the object to be tested, and the reflected light from the object to be tested 30 passes through the returning polarizing beam splitter 丨5 to amplify the surface topography of the object to be tested 30; quarter wave A wave plate 16 is interposed between the polarizing beam splitter 15 and the second objective lens 17. The light beam from the polarizing beam splitter 15 passes through the wave plate 16 and becomes a circularly polarized light. The second lens 18 is used. The reflected light returned from the polarizing beam splitter 15 passes; a Rotation Stage 19, which is an angle-variable platform; and an angle sensor 20 disposed on the rotating platform 19, using rotation The platform 19 rotates the 201237359 angle sensor 20 to make the incident angle of the reflected light from the third lens 18 The sensor 2 does multiple reflections' and controls the reflected light to be slightly angled near the critical angle. In one embodiment of the invention, the angle sensor 20 is a coated parallelogram ( Parallelogram Prism) 'The number of reflections is twice; and the matrix type photo sensor 21 is used for taking and extracting the light intensity change of the reflected light of the object to be tested, and transmitting it to a computer 22 for processing to calculate the object to be tested. 3 〇 surface topography; a specific embodiment of the present invention, the matrix type photo sensor 21 is a Charge-Coupled Device (CCD). 3.2 Measurement results The experimental framework of the present invention is shown in Fig. 6. The He-Ne laser is used as the light source 10, and passes through a light blocker (IS0lat0r) u to prevent the reflected light of the optical system from returning to the laser. Internally, the beam expander (Beam Expander) 12 is expanded into a parallel beam, and then passes through the polarizing plate 14 (P (0°)) to become p-polarized light, when the beam penetrates the second lens 14 and the polarized beam splitter (PBS). After 15, then the quarter-wavelength wave #16 becomes the gj pole, and then the sample of the object to be measured 30 to be measured (the present invention is mainly based on the grating and the /2 standard film) after the beam is irradiated The second objective lens 17, the polarizing beam splitter 15, the third lens μ, and the angle sensor 20 are imaged on the charge coupler (10) 21 . The matrix photo sensor 18 (ie, the charge coupler 18 (CCD)) is directly connected to the computer 4 by a universal serial bus connection (USB), and finally the image taken by the CCD can be displayed by the software on the computer 22. It is mainly the light-and-change diagram of each shot-to-gray scale when it is different from the total reflection (0.) and the critical oscillating (5· _.), respectively. 2 and I. , and then use Matlab software for analysis, the light intensity change picture taken by it is private 2 = L / I. 2 Go down and find the reflectance after the second reflection. Reuse the formula (^) i S ] 12 201237359 to get the height change, and finally draw the 3D wheel _. However, the angle sensor 2G (ie, Parallelogram Prism) is matched with the rotating platform (R〇tati〇n along the feeding (10), and the rotating platform 19 is used to control the angle sensor 20 to rotate back and forth to the total reflection angle and the criticality. Between the angles, the system error amount is first measured by the standard of 乂/2〇, and the roughness value measured by the atomic force microscope (as the reference value) is subtracted from the surface roughness value measured by the experimental structure of the present invention. After estimating the systematic error of the experimental framework of the present invention, the system error is corrected by using equations (12) and (^). 3. The measurement result of the non-penetrating test object (λ/2〇 standard film) will be The matrix photo sensor 21 is a charge coupler (Charge_c〇upled)
Device ’ CCD)於全反射以及臨界角附近所拍攝的光強度變化之灰階圖 形，得到/。,以及/。2，即可推得反射率進而求出高度變化。圖7為 Λ/20標準片之三維圖形。 3. 2. 2穿透型待測物之量測結果(Grating光拇） 清配合參看圖6、8及9所示，同樣地利用矩陣式光感測器21(即 電荷耦合器(Charge-Coupled Device ’ CCD))拍攝時將其角度感測器 20(即平行四邊形棱鏡(Parallel〇gram卩由爪))轉至全反射角以及臨界角 附近後’利用CCD各拍攝-張光強度變化之灰階圖形，制/〇ι以及 心’即可推得反射率仏進而求出高度變化，如圖8為施顏光 柵之一維圖形’而圖9為圖8所量測y軸於150/zm距離之曲線圖。 其拍攝别需難制物3〇表面至第二物鏡n之焦平面上，並使cCD 影像清晰’最好的調整方法是先於臨界角時作調整，由於臨界角附近 時反射率的變化較為靈敏，所以可以很明顯地得到較高對比度的變化[。] 201237359 3.2.2系統誤差修正 本發明系'統誤差修正係以Λ / 20標準片作討論，且利用原子力顯微 _之_度作為標準絲，織轉更接近所量測 之Μ值。因為本發明^架構所測得之粗糙度為APM所測得之粗糙度之 a倍’因此我們令本發明架構值為(12)式且⑽式之^倍如⑽ 式所示，此a值為修正之倍率。最後利用maTLAB求粗縫度時，只 要令本架構之原m值為(w/a)，經過修正之系統誤差後，本發明架構 所測得之〜（Average Roughnessm〜（R〇〇t 他肪 Square)[8]將跟 AFM所測得之〜與\相同。 M =Device ' CCD ) is a gray scale pattern of the change in light intensity taken near the total reflection and the critical angle, resulting in /. ,as well as/. 2, the reflectance can be derived to determine the height change. Figure 7 shows the three-dimensional graphics of the Λ/20 standard. 3. 2. 2 Measurement results of the penetrating test object (Grating light) Clear matching Referring to Figures 6, 8 and 9, the same use of the matrix photo sensor 21 (ie charge coupler (Charge- Coupled Device ' CCD)) When shooting the angle sensor 20 (ie parallelogram prism (Parallel〇gram卩 from the claw)) to the total reflection angle and the critical angle, 'taken by CCD - the intensity of the sheet light changes Gray-scale graphics, system / 〇ι and heart ' can be used to derive the reflectivity 仏 and then determine the height change, as shown in Figure 8 is a one-dimensional graph of the image grating' and Figure 9 is the measured y-axis of Figure 15 Zm distance curve. The shooting does not require the difficult surface of the object 3 to the focal plane of the second objective lens n, and make the cCD image clear. The best adjustment method is to adjust the angle before the critical angle, because the reflectivity changes near the critical angle. Sensitive, so you can clearly get a higher contrast change [. 201237359 3.2.2 System Error Correction The present invention is based on the Λ / 20 standard film, and uses the atomic force microscopy as the standard wire, and the weaving is closer to the measured Μ value. Since the roughness measured by the present invention is a times the roughness measured by APM', we have the inventive value of (12) and the formula (10) is as shown in (10), the value of a To correct the magnification. Finally, when using maTLAB to obtain the coarse seam degree, as long as the original m value of the structure is (w/a), after the corrected system error, the structure of the invention is measured ~ (Average Roughnessm~(R〇〇t Square)[8] will be the same as the \ measured by AFM. M =
\ Αχ , M' = -\ Αχ , M' = -
Mr = αΜ (12)(13) 肆·結論 本發明技絲縣光學反射式、非破壞性、大面積、可快速檢驗 之表面輪雜祕度制。若將物鏡之放大倍料何獲取如三維顯 微鏡的功效。不須掃瞄或斷層或影像堆疊，可直接求得三維之影像。 更重要的是本發明所研發出的架構，可以作大面積的制，因此不但 能節省時f植本外，亦騎破壞待測物絲面，林是轉透式或是 穿透式的待測物，只要有足夠的反射光強度，都可以做量測。本發明 技術屬-種光學式紐型三絲面形貌與顯微量職置。本發明可作 為瑕疵、缺陷、表面分析或粗糙度等量測。本發明可作為薄臈厚度量 測。本發明可作為透a賊非透明材料表面量測。本發明屬—種類似光 學系統(如顯微鏡)二維影像加上表面高度資訊所構成的三維電子兴 201237359 像。本發蚊-種_光贿歧射特換絲面高度的—種量測裝 置。本發明乃是-翻舰陣式域·(例如1躲合獅或 & # Λ &屬氧化物半導體（c_ementary Meta卜Oxide-Semiconductor，C0MS)作為擷取待測物影像，直每點 上光強度作為量測表面高度的資訊。本發明是包含_表面高度變化 所造成的反射光線角度變化的-種裝置。本發明是包含糊光線角度 偏向造成的反鮮絲酸改變的—雜置。本發明之反射率範圍^ 〇〜1.〇 ’角度範圍為土〇]〇度；表面高度量測範圍為〇. inra〜imm，縱向解 析度為0· 1簡〜10_ ;橫向解析度為_，卵；放大倍率可為 〇. 1〜200倍。本發明所使狀角度感測妓—種反射率或光強度對入射 角變化之轉換，可以為稜鏡或有細之元件。本發明所使用之 單一波長光源或採用雷射。 為 以上所述’僅為本發明之一可行實施例，並非用以限定本發明之 專利範園，凡舉依據下列請求項所述之内容、特徵以及其精神而為之 其他變化的等效實施，皆應包含於本發明之專利範圍内。本發明除上 述優點外，並深具產業之彻性，可有效改善習麟產生之缺失:、而 且所具體界定於請求項之雜’未見於醜物品，故而具實用性與進 步性，已符合新型專利要件，爰依法具文提出申請，謹請釣局依法 核予專利，以維護本申請人合法之權益。 【圖式簡單說明】 圖1為本發明光線於平行四邊形稜鏡内作兩次反射示意圖； 圖2為本發明兩次§偏光反射之反射率對外角的變化模擬圖； 圖3為本發明平行四邊形稜鏡之反射率實際量測曲線圖； 圖4為本發明架構之成像系統示意圖； 15 201237359 圖5為本發明待測物之幾何高度示意圖； 圖6為本發明實驗架構圖； 圖7為本發明又/20標準片之量測3D圖（包含系統誤差）； 圖8為本發明201ines/mm光栅之量測3D圖（包含系統誤差）；及 圖9為本發明取出圖8之乂轴於150 _距離之曲線圖（包含系統誤 差）。 附件：參考文獻。 【主要元件符號說明】 光源10 光阻隔器11 擴束器12 第一物鏡120 第一透鏡121 銷孔122 偏極板13 第二透鏡14 偏極分光鏡15 四分之一波片16 第二物鏡17 第三透鏡18 旋轉平台19 角度感測器20 平行四邊形棱鏡200 矩陣式光感測器21 電腦22 待測物30 s] 16Mr = αΜ (12)(13) 肆·Conclusion The surface of the invention is optically reflective, non-destructive, large-area, and can be quickly inspected. If the magnification of the objective lens is obtained, the effect of the three-dimensional microscope is obtained. Three-dimensional images can be directly obtained without scanning or faulting or image stacking. What's more important is that the structure developed by the present invention can be made into a large-area system, so that not only can the plant be saved, but also the surface of the object to be tested can be damaged, and the forest is transflected or transmissive. The measurement object can be measured as long as there is sufficient reflected light intensity. The invention belongs to the invention of an optical type three-filament surface topography and a microscopic amount. The invention can be used for measurements such as defects, defects, surface analysis or roughness. The invention can be used as a thin gauge thickness measurement. The invention can be used as a surface measurement of non-transparent materials. The invention belongs to a three-dimensional electronic 201273359 image which is composed of a two-dimensional image similar to an optical system (such as a microscope) and surface height information. This type of measuring device is a type of measuring device for the height of the silk surface. The present invention is a versatile array domain (for example, a lion or a &#; Λ & is an oxide semiconductor (c_ementary Meta Bu-Oxide-Semiconductor, C0MS) as an image of the object to be tested, at every point Light intensity is used as information for measuring the height of the surface. The present invention is a device comprising a change in the angle of reflected light caused by a change in the height of the surface. The present invention is a miscible change containing the change in the angle of the ray of the paste. The reflectance range of the present invention ^ 〇 ~1. 〇 'angle range is soil 〇 〇 ;; surface height measurement range is 〇. inra~imm, longitudinal resolution is 0 · 1 simple ~ 10_; lateral resolution is _ , the egg; the magnification may be 1. 1~200 times. The angle of the sensing of the present invention is a conversion of the reflectance or the light intensity to the change of the incident angle, which may be a 稜鏡 or a thin component. The use of a single-wavelength light source or the use of a laser. The above description is merely one of the possible embodiments of the present invention, and is not intended to limit the scope of the invention, which is based on the content, features, and Spiritual and other changes Equivalent implementations are all included in the scope of the patent of the present invention. In addition to the above advantages, the present invention has deep industrial rigor and can effectively improve the lack of Xilin production: and is specifically defined in the request item. It has not been found in ugly items, so it is practical and progressive. It has already met the requirements of new patents. It is required to file an application according to law. The fishing bureau is required to approve patents in accordance with the law to protect the legitimate rights and interests of this applicant. [Simplified description] 1 is a schematic diagram showing two reflections of light in a parallelogram crucible according to the present invention; FIG. 2 is a simulation diagram of a change in reflectance of a double-sided polarized reflection according to the present invention; FIG. 3 is a reflectance of a parallelogram of the present invention. Figure 4 is a schematic diagram of an imaging system of the present invention; 15 201237359 Figure 5 is a schematic diagram showing the geometric height of the object to be tested according to the present invention; Figure 6 is an experimental architecture diagram of the present invention; 3D map of the slice (including system error); FIG. 8 is a 3D diagram (including system error) of the 201ines/mm grating of the present invention; and FIG. 9 is the first axis of the present invention taken out at 150 _ distance Line diagram (including system error) Accessories: References [Main component symbol description] Light source 10 Light blocker 11 Beam expander 12 First objective lens 120 First lens 121 Pin hole 122 Polar plate 13 Second lens 14 Polar Beam splitter 15 quarter wave plate 16 second objective lens 17 third lens 18 rotating platform 19 angle sensor 20 parallelogram prism 200 matrix light sensor 21 computer 22 object to be tested 30 s] 16