1226435 政、發明說明： 一、 發明所屬之技術領域 本發明係關於一種光學式金屬辨識裝置及方法，特別係 關於一種可根據金屬對窄頻光之反射率差異辨識該金屬種 類之光學式金屬辨識裝置及方法。 二、 先前技術 目w印刷電路板之缺陷檢測主要係採用反射式白光影像 檢測法及反射式螢光影像檢測法二種。反射式白光影像檢 測法係以白光（White Light)照射在印刷電路板上，並利用 光感應彳貞測器擷取印刷電路板之金屬導線的反射光訊號， 再經由影像處理該反射光訊號而成為反射影像。而螢光影 像檢測法則是利用一雷射光激發該印刷電路板之非金屬導 線部分的有機螢光基材，配合光學濾光片將激發雷射光濾 除’再以影像感測器擷取由雷射光激發之螢光（Laser Induced FlUOrescence，UF)影像。該影像感測器所擷取的 灰階影像先經由二值化處理後，再經過影像辨識軟體之辨 識而獲得電路缺陷資訊。 習知之螢光檢測法雖可取得有機螢光基材的螢光訊號， 但由於金屬線材並無法產生螢光訊號，因此螢光檢測法無 法辨識金屬線材。反射式白光檢測法雖然可擷取金屬線材 的反射光訊號，但由於不同金屬對於白光的光學反射率差 異性並不大，因而並無法有效地辨識不同金屬線材。 三、 發明内容 本發明之主要目的係提供一種光學式金屬辨識裝置及方1226435 Description of policies and inventions: 1. Field of the invention The present invention relates to an optical metal identification device and method, and more particularly to an optical metal identification that can identify the type of metal based on the difference in reflectance between the metal and narrowband light. Device and method. 2. Prior art The defect detection of printed circuit boards mainly uses two methods: reflective white light image detection method and reflective fluorescent image detection method. The reflection type white light image detection method uses white light to illuminate a printed circuit board, and uses a light-sensing sensor to capture the reflected light signal of the metal wire of the printed circuit board, and then processes the reflected light signal through the image. It becomes a reflected image. The fluorescent image detection method uses a laser light to excite the organic fluorescent substrate of the non-metallic wire portion of the printed circuit board, and the optical laser filter is used to filter the excited laser light, and then the image sensor is used to capture the laser light. Laser Induced FlUOrescence (UF) image. The grayscale image captured by the image sensor is first binarized and then recognized by the image recognition software to obtain circuit defect information. Although the conventional fluorescent detection method can obtain the fluorescent signal of the organic fluorescent substrate, since the metal wire cannot generate the fluorescent signal, the fluorescent detection method cannot identify the metal wire. Although the reflective white light detection method can capture the reflected light signal of metal wires, it is not effective to distinguish different metal wires because the difference in optical reflectance of different metals to white light is not large. 3. Summary of the Invention The main purpose of the present invention is to provide an optical metal identification device and method.
H:\HU\HYG\ 丨:硏院通测中心\PD0065\PD0065.DOC 1226435 法’其可根據金屬對-窄頻光之反射率差異辨識該金屬之 種類。 為達成上述目的，本發明提供一種光學式金屬辨識裝 置，其包含一可承載一待測樣品之載台、一可產生至少一 窄頻光之光源模組、一用以聚焦該窄頻光於該待測樣品上 之透鏡組、-用以擷取該窄㈣經該待測樣品反射後之反 射光強度之光感測器以及-用以辨識該待測樣品所包含之 金屬種類之處理單元。 該處理單元可包含-影像辨識軟體，其根據該窄頻光經 該待測樣品反射後之反射光強度辨識該待測樣品所包含之 金屬種類。該光源模組可為—雷射…發光二極體或以一 白光光源加上-適當濾光片構成’其中該白光光源可為水 銀燈、鹵素燈、金屬齒辛擦武 哥®畜澄或巩虱燈等。該窄頻光之頻寬 較佳地係小於20nm，最佳地為小於5nm。 本發明之光學式金屬辨識方法包含以—第—人射角照射 該第-窄頻光於該待測樣品上’並祿取該第一窄頻光於該 第一入射角的反射光強度。德 反 < 设根據該第一窄頻光的反 射光強度辨識該待測樣品所包含之金屬種類。 本叙明之光學式金屬辨識方法亦可包含以一第二入射角 照射該第一窄頻光於該待測樣品，並擷取該第一窄頻光於 該第二入射角經該待測樣品反射後之反射光強度。之後， 根據該第一窄頻光於該第—入射角與該第二入射角之反射 光強度之差值辨識該待測樣品所包含之金屬種類。 本發明之光學式金屬辨識方法亦可包含以該第—入射角H: \ HU \ HYG \ 丨: Puyuan General Testing Center \ PD0065 \ PD0065.DOC 1226435 method, which can identify the type of metal based on the difference in reflectance between metal pair and narrow-band light. To achieve the above object, the present invention provides an optical metal identification device, which includes a stage capable of carrying a sample to be measured, a light source module capable of generating at least one narrow-band light, and a light-focusing module for focusing the narrow-band light on A lens group on the sample to be tested, a light sensor to capture the reflected light intensity of the narrow beam reflected by the sample to be tested, and a processing unit to identify the type of metal contained in the sample to be tested . The processing unit may include image recognition software that identifies the type of metal contained in the test sample based on the reflected light intensity of the narrow-band light reflected by the test sample. The light source module can be-a laser ... a light emitting diode or a white light source plus a suitable filter. 'The white light source can be a mercury lamp, a halogen lamp, a metal-toothed scorching wuge® animal or scorpion lamp. Wait. The bandwidth of the narrow-band light is preferably less than 20 nm, and most preferably less than 5 nm. The optical metal identification method of the present invention includes irradiating the first narrow-band light on the sample to be measured at a first incident angle and taking the reflected light intensity of the first narrow-band light at the first incident angle. De < Suppose that the type of metal contained in the test sample is identified based on the reflected light intensity of the first narrow-band light. The optical metal identification method described herein may further include irradiating the first narrow-band light on the sample under test at a second incident angle, and capturing the first narrow-band light at the second angle of incidence through the sample under test Intensity of reflected light after reflection. Then, the type of metal contained in the test sample is identified based on the difference between the reflected light intensity of the first narrow-band light at the first incident angle and the second incident angle. The optical metal identification method of the present invention may also include the first incident angle.
H:\HU\HYG\X 硏院遣测中心\PD0065\PD0065.DOC 1226435 、、第一乍頻光於該待測樣品上，並擷取該第二窄頻光 於σ亥第一入射角的反射光強度。之後，根據該第一窄頻光 、射光強度與該第二窄頻光的反射光強度之差值辨識該 待測樣品所包含之金屬種類。 相車乂於習知技藝，本發明之光學式金屬辨識裝置及方法 係利用金屬對窄頻光之反射率差異來辨識金屬之種類。由 於金屬對一窄頻光之反射率差異遠大於對白光之反射率差 異口此本發明應用於印刷電路上時可有效且正確地提昇 印刷電路板之金屬線材之辨識能力。 四、實施方式 圖1係本發明之光學式金屬辨識裝置1〇之示意圖。如圖 1所不，该光學式金屬辨識裝置丨〇包含一可承載一待測樣 品12之載台14、一可產生至少一窄頻光22之光源模組 20、一用以聚焦該窄頻光22於該待測樣品12上之透鏡組 24、一用以擷取該窄頻光22經該待測樣品}2反射後之反 射光強度之光感測器30以及一用以辨識該待測樣品丨2所 包含之金屬種類之處理單元4〇。該光源模組2〇產生之窄 頻光22經由該透鏡組24聚焦後可照射於該待測樣品12 之一預定區域上。該窄頻光22之頻寬較佳地係小於2〇nm， 最佳地為小於5nm。 5亥光源模組20可為一雷射、一發光二極體或以一白光光 源加上一適當濾光片構成，其中該白光光源可為水銀燈、 鹵素燈、金屬鹵素燈或氙氣燈等。該光源模組2〇最佳地為 雷射。該透鏡組24可由一聚焦透鏡組或一柱型透鏡組構H: \ HU \ HYG \ X 硏 院 测测 中心 \ PD0065 \ PD0065.DOC 1226435, the first light of the first frequency on the sample to be measured, and capture the second narrow-frequency light at the first incidence angle of σH Reflected light intensity. Then, the type of the metal contained in the test sample is identified based on the difference between the first narrow-band light, the intensity of the emitted light, and the reflected light intensity of the second narrow-band light. The photo car is familiar with the conventional art. The optical metal identification device and method of the present invention use the difference in reflectance of metal to narrow frequency light to identify the type of metal. Since the difference in reflectivity of a metal to a narrow-band light is much larger than the difference in reflectance to a white light, the present invention can effectively and correctly improve the discrimination ability of a metal wire of a printed circuit board when applied to a printed circuit. 4. Embodiment Figure 1 is a schematic diagram of an optical metal identification device 10 of the present invention. As shown in FIG. 1, the optical metal identification device includes a stage 14 capable of carrying a sample 12 to be measured, a light source module 20 capable of generating at least one narrow-band light 22, and a narrow-band light source for focusing. A lens group 24 of the light 22 on the sample 12 to be tested, a light sensor 30 for capturing the reflected light intensity of the narrowband light 22 reflected by the sample to be measured 2 and a light sensor 30 for identifying the sample to be tested The processing unit 40 of the metal type contained in the test sample 2. The narrow-band light 22 generated by the light source module 20 can be focused on a predetermined area of the sample 12 to be measured after being focused by the lens group 24. The bandwidth of the narrow-band light 22 is preferably less than 20 nm, and most preferably less than 5 nm. The Haihai light source module 20 may be a laser, a light emitting diode, or a white light source plus an appropriate filter. The white light source may be a mercury lamp, a halogen lamp, a metal halide lamp, or a xenon lamp. The light source module 20 is optimally a laser. The lens group 24 may be constituted by a focusing lens group or a cylindrical lens group.
H:\HU\HYG\r.硏院 Vft 测中心\PD0065\PD0065.DOC 1226435 成，用以將該光源模組20產生之光束聚焦於該待測樣品 12上。該光感測器30可為一電荷耦合感測元件，其可感 測該窄頻光22經該待測樣品12上預定區域反射後之反射 光強度，並據以產生該預定區域之反射光強度分佈影像。 ，處理單元40可包含-影像賴軟體，其可根據該光感測 為30產生之反射光強度分佈影像辨識金屬之種類。 圖2例不數種金屬之反射率與波長之關係。如圖2所示， 金與鋁對波長為400 nm之窄頻光的反射率分別為36·4%及 92.6%。若將該光源模組2〇產生之窄頻光22之中心波長設 定為400 nm，並設定照射該待測樣品12之光強度為a，Ζ 金與紹對該窄頻光22之反射率分別為4與^，則該窄頻 光22經該待測樣品12上之金與鋁反射後之反射光強度分 別為ft與尽％。若設定一門檻光強度為巧％與尽％之 平均值，並將反射光強度高於該門根光強度者判定為無， 且將反射光強度低於該門檻光強度者判定為金。如此，該 處理單元40即可根據該光感測器3〇產生之反射光強度分 佈影像辨識該預定區域内金與鋁之分佈情形。亦即，本發 明之光學式金屬辨識裝置1〇可根據該待測樣品12之金屬 對該窄頻光22之反射率的差值辨別該金屬之種類。 如上所述，應用本發明於辨識金與鋁時，較佳之窄頻光 之中心波長為400 nm。唯，若應用本發明之光學辨識方法 於辨識其它金屬時，則可依據待辨識金屬之反射率曲線， 選擇一個以上的適當窄頻光波長讓不同金屬的反射率具有 最大的差異，其中窄頻光之波長範圍可在紫外光區（1卯至H: \ HU \ HYG \ r. Puyuan Vft Test Center \ PD0065 \ PD0065.DOC 1226435 is used to focus the light beam generated by the light source module 20 on the sample 12 to be tested. The light sensor 30 can be a charge-coupled sensing element, which can sense the intensity of the reflected light after the narrow-band light 22 is reflected by a predetermined area on the sample 12 to be measured, and generate the reflected light of the predetermined area accordingly. Image of intensity distribution. The processing unit 40 may include an image-relying software which can identify the type of metal based on the reflected light intensity distribution image generated by the light sensing 30. Figure 2 illustrates the relationship between the reflectance and wavelength of countless metals. As shown in Figure 2, the reflectivity of gold and aluminum to narrow-band light with a wavelength of 400 nm is 36.4% and 92.6%, respectively. If the center wavelength of the narrow-band light 22 generated by the light source module 20 is set to 400 nm, and the intensity of the light irradiating the sample 12 to be measured is set to a, the reflectances of Z and Sha respectively In the case of 4 and ^, the reflected light intensities of the narrow-band light 22 after being reflected by the gold and aluminum on the test sample 12 are ft and%, respectively. If a threshold light intensity is set to be the average of Q% and X%, and the reflected light intensity is higher than the door root light intensity, it is judged as none, and the reflected light intensity is lower than the threshold light intensity as gold. In this way, the processing unit 40 can identify the distribution of gold and aluminum in the predetermined area based on the reflected light intensity distribution image generated by the light sensor 30. That is, the optical metal identification device 10 of the present invention can discriminate the type of the metal based on the difference between the reflectance of the metal of the sample 12 to be measured and the narrowband light 22. As described above, when the present invention is applied to identify gold and aluminum, the preferred narrow wavelength light has a center wavelength of 400 nm. However, if the optical identification method of the present invention is used to identify other metals, according to the reflectivity curve of the metal to be identified, one or more appropriate narrow-band light wavelengths can be selected to make the reflectance of different metals have the largest difference. The wavelength of light can be in the ultraviolet region (1 卯 to
H:\HU\HYG\ 丨::硏院\撒測中心\PD0065\PD0065.DOC 1226435 380 nm)、可見光區（38〇至78〇 nm)和近紅外光（78〇—乃⑻ nm)。簡言之，本發明係依據金屬對不同波長的反射率曲 線，選擇一個以上的適當窄頻光波長讓不同金屬的反射率 具有最大的差異。 弗列斯涅爾方程式（Fresnel equation)可表示如下： Γ丄=& = fe/cosW-fa/cosA) (72 /cos 0 ) + (% /cos Γ" = i = "2 c〇s c〇s <9,·H: \ HU \ HYG \ 丨 :: Puyuan \ Scattering Test Center \ PD0065 \ PD0065.DOC 1226435 380 nm), visible light region (38-78 nm), and near-infrared light (78-Nano-nm). In short, the present invention selects more than one appropriate narrow-band light wavelength based on the reflectivity curves of metals to different wavelengths so that the reflectance of different metals has the largest difference. The Fresnel equation can be expressed as follows: Γ 丄 = & = fe / cosW-fa / cosA) (72 / cos 0) + (% / cos Γ " = i = " 2 c〇sc 〇s < 9, ·
Ei〇 72 c〇s^ cos (9. 其中’ Γ±、Γ〃分別為垂直極化與平行極化的反射係數；％、 仏分別為介質的本質阻抗；6、0則分別為入射光角度與折 射光角度。由弗列斯涅爾方程式可知，金屬之反射率隨著 入射角度而改變。本發明之光學式金屬辨識裝置1〇亦可利 用此一物理特性辨識金屬。首先以一第一入射角照射該窄 頻光22於該待測樣品12上，並擷取該窄頻光22於該第一 入射角之反射光強度。之後，以一第二入射角照射該窄頻 光22於該待測樣品12上，並擷取該窄頻光22於該第二入 射角之反射光強度。由於金屬之反射率隨著入射角度而改 變’因此該處理單元4〇可根據該窄頻光22於該第一入射 角與該第二入射角之反射光強度之差值辨識該待測樣品 12所包含之金屬種類。 本發明亦可使用二種不同波長之窄頻光來辨識金屬之種 類。一般而言，一印刷電路板包含由銅構成之金屬導線及 由金構成之接觸墊。較佳地，本發明可使用中心波長分別 H:\HU\HYGU:硏院 V麗测中6\PD0065\PD0065.D〇C -10- 1226435 為400 nm之第一窄頻光及中心波長為6〇〇 之第二窄頻 光。若該第一窄頻光及該第二窄頻光於該印刷電路板之照 射光強度分別設定為及A。。，則該印刷電路板之銅對該 第一窄頻光及該第二窄頻光之反射光強度之差值為 - A。。X - £4。。X i?c_，其中 a·、t，4。。係分別為銅對該 第一窄頻光及該第二窄頻光之反射率。同理，該印刷電路 板之金對該第一窄頻光及該第二窄頻光之反射光強度之差 值為·· 仏。α,6。。一五4。。％，4。。，其中I。。、及我係分別為 金對該第一窄頻光及該第二窄頻光之反射率。 應用本發明於該印刷電路板之金屬辨識時，可先以該第 一窄頻光照射該印刷電路板，並由該光感測器3〇擷取該第 一窄頻光經該印刷電路板反射後之反射光強度。之後，以 第一窄頻光照射該印刷電路板，並由該光感測器3 〇擷取該 第一乍頻光經該印刷電路板反射後之反射光強度3若制定 一門檻光強度為與之平均值，則該處理單元4〇即可 將反射光強度南於該門檻光強度者判定為金，而將反射光 強度低於該門檻光強度者判定為銅。如此，該光學式金屬 辨識裝置10即可辨別該印刷電路板上金與銅之分佈區域。 相較於習知技藝，本發明之光學式金屬辨識裝置及方法 係利用金屬對窄頻光之反射率差異來辨識金屬之種類。由 於金屬對窄頻光之反射率差異遠大於對白光之反射率差 異’因此本發明應用於印刷電路上時可有效且正確地提昇 印刷電路板之金屬線材之辨識能力。 本發明之技術内谷及技術特點巳揭示如上，然而熟悉本 H:\HU\HYGU:硏院遺测中,C、\PD0065\PD0065.D〇C -11 - 1226435 項技術之人士仍可能基於本發明之教示及揭示而作種種不 背離本發明精神之替換及修飾。因此，本發明之保護範圍 應不限於實施例所揭示者，而應包括各種不背離本發明之 替換及修飾，並為以下之申請專利範圍所涵蓋。 五、 圖式簡要說明 圖1係本發明之光學式金屬辨識裝置之示意圖；以及 圖2例示數種金屬之反射率與波長的關係。 六、 元件符號說明 10光學式金屬辨識裝置12待測樣品 14載台 2 0光源模組 22第一窄頻光 24透鏡組 30光感測器 40處理單元 H:\HU\HYGV+1：硏院 \ 量测中6\PD0065\PD0065.DOC - 12-Ei〇72 c〇s ^ cos (9, where 'Γ ±, Γ〃 are the reflection coefficients of vertical polarization and parallel polarization, respectively;%, 仏 are the intrinsic impedance of the medium; 6, 0 are the angle of incident light, respectively And the angle of refracted light. It can be known from the Fresnel equation that the reflectivity of a metal changes with the angle of incidence. The optical metal identification device 10 of the present invention can also use this physical characteristic to identify a metal. First, a first The incident angle illuminates the narrow-band light 22 on the sample to be measured 12 and captures the reflected light intensity of the narrow-band light 22 at the first incident angle. Thereafter, the narrow-band light 22 is illuminated at a second incident angle on The reflected light intensity of the narrow-band light 22 at the second incident angle is captured on the sample 12 to be measured. Since the reflectivity of the metal changes with the incident angle, the processing unit 40 can use the narrow-band light 22 Identify the type of metal contained in the test sample 12 based on the difference between the reflected light intensity of the first incident angle and the second incident angle. The present invention can also use two kinds of narrow-frequency light with different wavelengths to identify the type of metal In general, a printed circuit board contains Metal wires made of copper and contact pads made of gold. Preferably, the present invention can use the central wavelengths H: \ HU \ HYGU: Puyuan V Lizhong 6 \ PD0065 \ PD0065.D〇C -10- 1226435 The first narrow-band light with a wavelength of 400 nm and the second narrow-band light with a center wavelength of 600. If the light intensity of the first narrow-band light and the second narrow-band light on the printed circuit board is set to and A ..., then the difference between the reflected light intensity of the copper of the printed circuit board to the first narrowband light and the second narrowband light is-A ... X-£ 4. X i? C_, where a ·, T, 4. are the reflectances of copper to the first narrowband light and the second narrowband light. Similarly, the gold of the printed circuit board is to the first narrowband light and the second narrowband light. The difference in the intensity of the reflected light of the frequency light is 仏 .α, 6..15..4.%, 4 .., where I .. and I are the first narrow frequency light and the gold respectively. Reflectivity of the second narrow-band light. When applying the present invention to the metal identification of the printed circuit board, the first narrow-band light may be irradiated to the printed circuit board first, and the first 30 A narrowband light beam The intensity of the reflected light after the printed circuit board reflects. After that, the printed circuit board is illuminated with the first narrow-band light, and the light sensor 30 captures the reflection of the first first-frequency light reflected by the printed circuit board. If the light intensity 3 sets a threshold light intensity as an average value, the processing unit 40 can determine that the reflected light intensity is lower than the threshold light intensity as gold, and the reflected light intensity is lower than the threshold light intensity. It is judged to be copper. In this way, the optical metal identification device 10 can distinguish the distribution area of gold and copper on the printed circuit board. Compared with the conventional art, the optical metal identification device and method of the present invention use a narrow metal pair. The reflectance difference of frequency light is used to identify the type of metal. Because the reflectance difference of metal to narrow frequency light is much larger than the reflectance difference of white light ', the invention can effectively and correctly improve the printed circuit board when applied to printed circuits. Identification of metal wires. The technical inner valley and technical characteristics of the present invention are disclosed as above. However, people familiar with this H: \ HU \ HYGU: 硏 Yuan Yuan test, C, \ PD0065 \ PD0065.D〇C -11-1226435 technology may still be based on The teachings and disclosure of the present invention are made in various substitutions and modifications without departing from the spirit of the present invention. Therefore, the protection scope of the present invention should not be limited to those disclosed in the embodiments, but should include various substitutions and modifications that do not depart from the present invention, and are covered by the following patent application scope. V. Brief Description of the Drawings Figure 1 is a schematic diagram of the optical metal identification device of the present invention; and Figure 2 illustrates the relationship between the reflectance and wavelength of several metals. VI. Symbol descriptions 10 Optical metal identification device 12 Sample to be tested 14 Stage 2 0 Light source module 22 First narrowband light 24 Lens group 30 Light sensor 40 Processing unit H: \ HU \ HYGV + 1: 硏Hospital \ Measuring 6 \ PD0065 \ PD0065.DOC-12-