TW201800720A - Optical system and method of surface and internal surface profilometry using the same - Google Patents

Abstract

The present invention provides an optical system and method for full-field high-speed object surface and internal (including bottom) surface profilometry using simultaneous multi-point optical scanning. In the optical system, at least one point light generated by an optical modulating unit is projected onto at least one position on an object surface and an object internal (including bottom) surface. The reflected light from the object surface and internal (including bottom) surface are detected by an image acquiring device, thereby generating an optical diffractive image with at least one detected optical diffractive pattern respectively corresponding to the depth of the tested surface point. The detection method is utilized to perform a vertical scanning on each calibrated depth for establishing a database having a plurality of optical diffraction patterns being corresponding to each calibrated depth, respectively. A control and calculation unit is utilized in the optical system to perform normalization calculation of cross correlation between the measured diffraction pattern and a plurality of calibrated diffraction patterns for determining the measured depth of the tested object surface and internal (including bottom) surface position in one-shot imaging manner.

Description

Optical system and method for detecting three-dimensional shape of surface or internal light reflection interface of object using the same

The invention relates to a high-speed global optical system and detection technology by synchronous multi-point optical scanning, in particular to using a scattering pattern image to detect the three-dimensional shape of an object surface or an internal light reflection interface (including the back side of the object). An optical system, and a method for detecting a three-dimensional shape of an object surface or an internal light reflecting interface using the system.

The color confocal microscopy system is one of the methods for measuring the three-dimensional shape of the surface of an object. It can measure the height, line width, groove width and depth of the mechanical or semiconductor structure, and is therefore important for process improvement or yield detection. in accordance with. This technique was first proposed by Marvin Minsky in 1957. The principle of color confocal is to disperse incident light to form multiple detection lights with different depths of focus to form a measurement mechanism for optical vertical scanning. Using this method to detect the object to be tested, opticals of different depths can be obtained. The sliced image is filtered by a pinhole to remove the reflected light and the scattered light outside the focus area, and the focus surface information is retained, and the optical slice image obtained by different depths is reconstructed by the computer, that is, The three-dimensional image information of the object to be tested can be obtained.

A confocal wafer inspection system as disclosed in U.S. Patent No. 6,934,019, the light field of a light source being projected through a lens to focus at different focus positions. Because of the point light field, the light field having only one color transmitted through the reflected light field on the wafer to be tested can pass through the filter element through the reflection of the beam splitter 14. The surface height of different locations on the wafer to be tested is measured by moving the object to be tested or moving the optical structure. Although the aforementioned conventional technique can measure the surface height of the object to be tested, since the position of the focus is a point light field, the position of each detection is only a single point, so it is possible to measure the surface three-dimensionality of the entire object to be tested. The appearance is not only time consuming but also reduces the production efficiency of the process. In addition, since the reflected light field is a single color light, it can be directly analyzed by the spectrometer.

In addition, a confocal microscopy apparatus is disclosed in U.S. Patent No. 5,785,651. In this technique, the confocal microscopy device utilizes a polychromatic light generated by a light source through an achromatic collimator lens, and then forms a collimated light field without chromatic aberration and projects onto the Fresnel optical element. . After the Fresnel optical element, a scattered light field having different focusing points with different wavelengths is formed to detect the three-dimensional shape of the surface of the object to be tested. In this technique, similarly, the light field is modulated to a point that is focused at different positions with different wavelengths. Since each detection position is only a single point, it is necessary to be able to measure the entire object to be tested. The three-dimensional shape of the surface is not only time consuming but also reduces the production efficiency of the process. In addition, since the reflected light field is a single color light, it can be directly analyzed by the sensing element. In addition, the confocal detection sensor disclosed in U.S. Patent Application Publication No. 2004/0109170 also divides the light field into different wavelengths and respectively focuses on different focus positions.

Although the color confocal detection light can be exempted from the traditional vertical scanning because of the different depth of focus, because the movement of the vertical moving mechanism causes vibration problems caused by the measuring machine, there are still several parts to be solved:

First, in the case of multi-point simultaneous measurement, the noise of stray light and stray light between adjacent measurement points overlap and the noise of cross talk is easy to occur. In the conventional technology, in order to reduce the mutual interference of light. The problem is that pinholes are often placed in front of the spectrometer. However, this method can reduce the interference, but if it is applied to measure the two-dimensional surface topography of the object or requires high-resolution measurement, it needs to cooperate with the lateral movement. The mechanism will increase the time required for measurement and the influence of vibration on measurement accuracy. Therefore, the general method of confocal three-dimensional shape measurement is still applied in the field of single-point measurement.

Second, the spectrometer volume takes up space and the problem of analyzing the time-consuming spectrum. In the conventional technology, in order to reduce the three-dimensional shape of the surface, it is necessary to detect the spectrum of the light passing through the slit or the pinhole structure through a spectrometer, thereby finding the wavelength of the light corresponding to the detected position and its light intensity, and restoring the depth corresponding to the position. There are several ways to overcome the spectrometer. First, the spectrometer is bulky and takes up space. Furthermore, since the spectrometer reduces the spectral information time and does not have measurement efficiency, the main reason is that for each measurement site, the unfolded spectrum is a one-dimensional spectrum, if for a linear measurement position, the expansion It is a two-dimensional spectrum, which is unfavorable in both measurement speed and data calculation. Generally, the measurement speed is greatly reduced. It can be seen that each spectrometer performs one-time analysis, which is at most a linear measurement area. Therefore, if one surface is to be measured, it is necessary to scan through the lateral movement, so that not only the measurement time is increased. It is more likely that the lateral mechanical scanning motion causes vibration problems on the measuring machine, which affects the measurement accuracy.

At present, in optical micro-focusing measurement technology, the measurement signal is mainly obtained by mechanical or optical axial depth scanning by means of a single intensity or spectral information corresponding to the measured depth. Measure the light intensity or spectral information peak finding algorithm to find the depth position of the conjugate focal length. Since a single light intensity or spectral information is greatly affected by the noise of the photo sensor or the instability of the measurement system or the measurement environment variation, the depth information obtained by the measurement is likely to cause a large measurement error. The measurement accuracy is affected, and it cannot be further upgraded to a superior level. This is a key issue for the current optical microscopic confocal measurement method.

In summary, there is a need for an optical system and a three-dimensional topography detection method for the surface or internal light reflecting interface (including the back side of the object) using the system to solve the problems caused by the conventional technology.

The invention provides a method for detecting a three-dimensional shape of an optical system and an object surface or an internal light reflection interface, which utilizes a digital light modulation unit, such as a digital mirror device (DMD) or a liquid crystal liquid crystal element (liquid). Crystal on silicon, LCOS), generating a plurality of point sources with a specific distance, projecting onto the surface of the object or the internal light reflection interface, performing multi-channel optical scanning, and then measuring the optical diffraction pattern (or point distribution function, Point Spread Function (PSF)), by establishing a database of optical diffraction pattern samples about the depth, performing cross-correlation calculation on the optical diffraction pattern of at least one position on the object to be tested and the optical diffraction image sample in the database, Know the corresponding depth. Since the light modulation unit modulates the incident light into a plurality of point light sources to be projected onto the surface of the object to be tested, the noise of the cross talk due to the overlap of the light points can be greatly reduced, thereby improving the three-dimensional shape. The accuracy of the depth measurement of the appearance or measurement interface.

The invention provides a three-dimensional topography detecting method for an optical system and an object surface or an internal light reflecting interface, which can generate a one-dimensional or two-dimensional array point light source, position switching through a point light source and detection with an image capturing device Frame detection rate, three-dimensional topography detection of high-speed full-field object surface or internal light reflection interface, so that the optical system of the present invention can be quickly not only quaisi-single shot The full field surface contour information of the object to be tested is obtained, wherein the quasi-click measurement method is a global multi-point confocal measurement method, and is applied to an in-situ instantaneous measurement. In addition, the measurement method by clicking can also avoid the negative influence of the vibration generated by the vertical or lateral movement scanning on the measurement result, and the accuracy of the measurement result can be greatly improved.

In one embodiment, the present invention provides an optical system including a light source module, an objective lens, an image capturing device, and a control and calculation unit. The broadband light source module is configured to generate at least one light source. The objective lens is configured to project the at least one light source to at least one location of an object surface or an internal light reflecting interface. The image capturing device is configured to receive at least one reflector light reflected by the at least one location to generate a corresponding optical diffraction image having at least one optical diffraction pattern, each optical diffraction pattern being accurate Corresponding to a depth position of one of the surface or internal light reflecting interface of the object. The control and calculation unit calculates an optical diffraction pattern corresponding to each position in the optical diffraction image and a plurality of optical diffraction pattern samples to determine the depth of each position.

In another embodiment, the present invention provides a method for detecting a three-dimensional shape of an object surface or an internal light reflecting interface, comprising the steps of: firstly generating at least one light source; and then projecting the at least one light source to an object by an objective lens. At least one depth position of the surface of the object or the internal light reflecting interface; then, passing the at least one reflective object reflected by the at least one position through the objective lens; and then, guiding the at least one reflected object light to an image capturing device; Thereafter, the image capturing device senses the at least one reflector light to generate an optically diffracted image having at least one optical diffraction pattern corresponding to at least one depth position. Finally, the at least one optical diffraction pattern is respectively extracted from the optical diffraction image, and each optical diffraction pattern and a plurality of optical diffraction pattern samples are calculated to determine a surface reflection or a corresponding internal light reflection of each object. The interface measures the depth information of the location.

In an embodiment, wherein determining the depth of each location further comprises the following steps: First, a database is created having a plurality of optical diffraction pattern samples corresponding to different depths. Then, each of the captured optical diffraction patterns and the plurality of optical diffraction pattern samples corresponding to the position of the point source corresponding to the corresponding point source in the database are operated to generate a plurality of samples about the optical diffraction pattern. Normalized cross correlation (NCC) values. Finally, the depth corresponding to the optical diffraction pattern sample corresponding to the maximum of the plurality of normalized cross-correlation values is used as the measured depth corresponding to the position of the optical diffraction pattern. The establishing the database further comprises the steps of: providing a mirror surface having a first distance from the objective lens; and generating, by the at least one correction point light source, the dispersion through the object lens to the mirror surface; and reflecting from the specular surface After passing through the objective lens, the reflected light is guided to the image capturing device to generate a calibrated optical diffractive pattern for the first distance, which has at least one optical diffraction pattern sample. Corresponding to each of the correction point light sources; respectively extracting the at least one optical diffraction pattern sample about the first distance; changing the first distance to a second distance to further obtain another Correcting the optical diffraction image; and respectively extracting the at least one optical diffraction pattern sample for the second distance.

Various illustrative embodiments may be described more fully hereinafter with reference to the accompanying drawings. However, the inventive concept may be embodied in many different forms and should not be construed as being limited to the illustrative embodiments set forth herein. Rather, these exemplary embodiments are provided so that this invention will be in the In the drawings, the size and relative sizes of layers and regions may be exaggerated for clarity. Similar numbers always indicate similar components. The magnetic sensing device will be described below in conjunction with various embodiments, however, the following embodiments are not intended to limit the invention.

Please refer to FIG. 1A, which is a schematic diagram of an embodiment of an optical system of the present invention. The optical system 2 includes a light source module 20, an objective lens 21, an image capturing device 22, and a control and calculation unit 23. The broadband light source module 20 is configured to generate at least one light source. The at least one point of light source can be a single point source, a point source that presents a one-dimensional array, or a point source that presents a two-dimensional array. In an embodiment, the light source module 20 includes a light source 200, a collimating mirror set 201, and a light modulating unit 202. The light source 200 is used to generate an incident light 90. The incident light 90 of the present embodiment is a broadband light, for example, a white light source, but is not limited thereto. For example, laser light can also be implemented. The collimating lens group 201 receives the incident light 90, and then collimates the incident light to form a collimated light field 91, which is projected to the light modulating unit 202. The light modulating unit 202 can generate a pinhole filtering effect for modulating the collimated light field 91 into the at least one light source. The light modulation unit 202 can be selected as a digital mirror device (DMD) or a liquid crystal on silicon (LCOS).

The foregoing light modulation unit 202 has a plurality of light control switches 203, which can change the path of the light according to the control signal. For example, taking the DMD component shown in FIG. 2A as an example, the light modulating unit 202 has a plurality of reflective elements arranged in a two-dimensional array as the light control switch 203, and only four are represented in the figure. In general, the light control switch 203 has at least an on state and an off state. The light control switch 203 can change the angle of its rotation through the control of the telecommunication, thereby presenting an on or off state. When light is projected onto a plurality of light control switches 203, the path of the reflected light is determined according to the direction in which the light is deflected. In an embodiment, as shown in FIG. 2B, when the light control switch 203a is in the open state, the light reflecting element is deflected to an angle to reflect the light to the objective lens, and the light control switch 203b is in the off state. Its reflective elements reflect light to other places. In an embodiment, as shown in FIG. 2C, four light control switches 203a (2x2) may be used as a unit for generating a point light source, and a deflection direction of the light control switch 203b around the light control switch and a light control switch as a point light source. 203a has different deflection directions. The direction in which the light control switch 203a as the point light source is deflected can guide the reflected light to the objective lens. In addition, through such a multi-channel optical scanning method, when a plurality of point light sources are generated, due to the spatial distance D between them, the crosstalk problem of the confocal measurement signals between adjacent measurement points can be avoided. The LCOS is a light-transmitting unit that passes through the liquid crystal as light passing or failing, and can also be used as a light modulating unit of the present invention, which is well known to those skilled in the art, and will not be described herein.

Returning to FIG. 1A, a light splitting element 24 is provided between the objective lens 21 and the light modulating unit 202 for guiding at least a point of light from the light modulating unit 202 to the objective lens 21. The image capturing device 22 differs depending on the type of the objective lens. In an embodiment, if the objective lens 21 is an objective lens for general focusing, the image capturing device 22 may be a color or monochrome image capturing device for generating gray scale or multiple colors (light wavelength). An optical diffraction image in which the monochrome image capturing device can be a 12-bit monochrome CCD. It should be noted that the depth measurement range has a great relationship with the depth of focus of the objective lens. If we use a numerical aperture (NA), although we can increase the measurement slope, we choose a high NA. The objective lens reduces the depth of its focus. In order to balance the high NA value and not excessively reduce the range of the measurable depth distance, in another embodiment, as shown in FIG. 1B, the objective lens 21a may employ an axial chromatic objectives, such as a refractive type. Or a diffractive dispersion objective lens, and at the same time, the image capturing device 22a selects a color image capturing device, for example, a 12-bit color CCD. In an embodiment, the dispersion objective lens 21a divides each incident point light source into a plurality of color lights, and the present embodiment is represented by red light R, green light G, and blue light B. The focus points of each color light are separated by a distance D. In this way, it is ensured that a distinctive optical diffraction image can be captured on the color image capturing device to facilitate subsequent searching for the peak value of the corresponding normalized cross-correlation coefficient, thereby reconstructing the surface of the object or the internal light reflecting interface. Three-dimensional topography (including the back of the object).

Referring back to FIG. 1A, the objective lens 21 is used to project at least one collimated light field onto an object to be tested 8 disposed on the loading platform 25, and the object light 93 reflected by the object to be tested 8 follows the original light path. Returning to the objective lens 21, the light splitting by the spectroscopic element 23 is projected to the image capturing device 22. The image capturing device 22 receives the at least one reflector light 93 reflected by the at least one location to generate a corresponding optical diffraction image having an optical diffraction pattern corresponding to the at least one location.

In addition to the architecture of the dispersive objective lens of FIG. 1B, as shown in FIG. 1C, it is a schematic diagram of another embodiment of the architecture of the axial dispersive objective lens. The optical modulation unit 202 in this embodiment is different from the architecture of FIG. 1B. The effect of two pinhole filtering is generated, that is, the light source 200, for example, a white light source, generates incident light, and then is projected to the light modulating unit 202 through the beam splitting element 24, at which time the light modulating unit 202 plays the role of the first pinhole filtering. The incident light is modulated into at least one point of light source and then reflected to the collimating mirror group 201. The modulation mode can control the first region of the light modulation unit 202 to control the collimated broadband light into the at least one light source, and reflect to the dispersion objective lens 21a, and control the light modulation unit 202 to surround the light modulation unit 202. The second region around the periphery of the first region reflects the collimated incident light to a region different from the dispersive objective lens 21a.

The dispersion objective lens 21a projects the spot light source and then projects it onto the object to be tested 8. The reflector light reflected by the object to be tested 8 follows the original light path and is again projected to the light modulation unit 202. At this time, the light modulation unit 202 plays the role of the second pinhole filtering, that is, the light modulation unit 202 is controlled by the control signal. The first region reflects the reflector light to the beam splitting element 24, and controls the second region of the light modulating unit 202 that surrounds the periphery of the first region to direct the reflector light to a different from the beam splitting element 24. region. The spectroscopic element 24, which receives the reflected light, is guided to the focusing mirror group 26, and is focused and projected onto a color image capturing device 22a, such as a 12-bit color CCD, to produce an optically diffracted image. The architecture of the two-pinhole filtering through the light modulating unit 202 in FIG. 1C can increase the confocal effect and increase the axial resolution. It is to be noted that the objective lens of Fig. 1C is not limited to the dispersion objective lens 21a, and may be a general objective lens. If a general objective lens is used, the image capturing device 22a uses a monochrome image capturing device.

1A to 1C are embodiments for measuring the depth of the surface topography of an object, but the technique for reducing the topography of the surface to be measured by using the diffraction pattern according to the present invention is not limited to measuring the topography of the surface of the object. In another embodiment, the light source that is penetrating is projected onto the object to be tested, and the topography of the light reflecting interface (including the back side of the object) inside the object to be tested is measured. The inside can be the lower surface of the object to be tested in contact with the carrier, or the structural surface inside the object to be tested, such as a defect or a hollow structure, as shown in FIG. 1D, in the present embodiment, FIG. 1A The optical system architecture is used to illustrate that the incident light generated by the light source in the embodiment is a light source having transparency to the object to be tested. In an embodiment, for example, the light source is selected from infrared light, and the object to be tested is a germanium wafer. Under this structure, the object to be tested 8a is penetrated and reflected on the bottom surface of the object to be tested 8a to form the object light 93a, and then reflected through the objective lens 21 to form a shape corresponding to the bottom surface of the object to be tested on the image capturing device 22. Depth image of depth. In addition, in another embodiment, in the object to be tested 8a, if there is a crack or an inner layer hollow structure 83, the incident light is reflected on the crack surface or the hollow structure surface to form the object light 93b, and A diffraction pattern corresponding to the surface of the crack or the surface of the inner structure is produced on the image capturing device 22. It is to be noted that although FIG. 1D is illustrated by the optical system of FIG. 1A, the optical system is not limited in use, and the optical systems of FIGS. 1B and 1C can also be applied.

Next, the principle of the optical diffraction image of the present invention will be described. In optical, Fraunhofer diffraction, also known as far-field diffraction, is a kind of wave diffraction, when electromagnetic waves pass through a circular hole or slit. Occurs, resulting in a change in the size of the observed image. The cause is the far-field position of the observation point, and the diffraction wave passing outward through the circular hole has the property of a progressive plane wave. In the present invention, since the distance from the point light source generated by the light modulating unit 202 to the object to be tested 8 is much larger than the size of the point light source generated in the light modulating unit 202, for example, the size of the 2x2 light control switch, it is applicable to The principle of Lang and Fei diffraction.

…..(1) 其中z為點光源至光學繞射圖案平面S’的距離，A為光源振幅，a, b為代表空白區域矩形點光源的尺寸。當原點改成矩形點光源的中心位置時，進行積分的結果，可以得到如式(2)所示：

….. (2) 其中

如下式(3)所示：

….. (3) 其中

如下式(4)所示：

….. (4) 其中

代表沿著Y方向的光學繞射角度。 根據圖1A的架構，當待測物和光源的距離改變，不同聚失焦(in and out focus) 的光學繞射圖案會被影像擷取裝置所擷取，而該光學繞射圖案上每一個位置的強度分布則如式(5)所示：

….. (5)In an embodiment, since the array of light control switches 203 that generate point sources in the light modulating unit 202 can be regarded as a rectangular point source, it can be analyzed using the architecture shown in FIG. In Fig. 3, S represents a point source plane, wherein the hatched area represents the light control switch 203b that reflects light to another place, and the blank area represents the light control switch 203a that reflects light to the objective lens 21. Therefore, the blank area can be regarded as a rectangular point source. S' represents the plane of the optical diffraction pattern. According to the above structure, the light field u(x, y, z) of each position P on the optical diffraction pattern generated on the S' plane through the point source can be represented. As shown in equation (1):

..... (1) where z is the distance from the point source to the optical diffraction pattern plane S', A is the source amplitude, and a, b is the size of the rectangular point source representing the blank area. When the origin is changed to the center position of the rectangular point source, the result of integration can be obtained as shown in equation (2):

..... (2) where

As shown in the following formula (3):

..... (3) where

As shown in the following formula (4):

..... (4) where

Represents the optical diffraction angle along the Y direction. According to the architecture of FIG. 1A, when the distance between the object to be tested and the light source changes, different in and out focus optical diffraction patterns are captured by the image capturing device, and each of the optical diffraction patterns The intensity distribution of the position is as shown in equation (5):

..... (5)

Returning to FIG. 1A, the control and calculation unit 23 can determine the position of each point source on the surface of the object to be tested 8 according to the optical diffraction image of the object 8 and the optical diffraction pattern sample in the database. depth. That is, the data library electrically connected to the control and calculation unit 23 stores a plurality of optical diffraction pattern samples corresponding to the positions of the plurality of point light sources, and the optical diffraction images have optical diffraction corresponding to each of the point light sources. a pattern, so each control and calculation unit 23 captures each optical diffraction pattern and normalizes cross-correlation (NCC) calculations with a plurality of optical diffraction pattern samples corresponding to the same point source position in the database. The calculation unit 23 can determine the depth of the surface of the object to be tested corresponding to the position of the point source based on the result of the calculation.

Further, the control and calculation unit 23 changes the position of the point light source, and each time it is changed, another optical diffraction image is obtained, and the optical diffraction pattern corresponding to the position of the different point light sources is also obtained, and after the extraction, the The plurality of optical diffraction pattern samples corresponding to the position of the point source are calculated, and the depth information of the position of the point source corresponding to the position on the surface of the object to be tested can be obtained. After the optical diffraction pattern on a plurality of positions on the surface of the object or the internal light reflection interface is collected, and the depth information is calculated, the three-dimensional shape of the surface of the object to be tested can be reconstructed. It should be noted that whether to change the position of the point source by using the light modulation unit 202 may be determined according to the resolution of the reconstructed image. If the number of the point sources is sufficient, and the resolution required for the detection is sufficient, the unimorphism may be used. Capture the image. In addition, for the control mode of switching the position of the point light source multiple times, since the position of the point light source is changed by controlling the light control switch, the detection speed of the image capturing device is larger than that of the image capturing device, so the image capturing device detects The faster the rate, the faster the scanning of the surface of the object or the internal light reflecting interface can be done. Through such a scanning method, the mechanism of the horizontal scanning can be avoided, thereby reducing the interference of the vibration in the measurement, and improving the speed and accuracy of the surface three-dimensional shape detection.

Next, a method for measuring the three-dimensional shape of the surface or internal light reflecting interface of the present invention will be described. First, step 30 is provided to provide the optical system 2 as shown in FIG. 1, and a database having a plurality of optical diffraction pattern samples is provided. The plurality of optical diffraction pattern samples correspond to sample information of a light wavelength, and the embodiment is white light. The database for establishing the optical diffraction pattern sample includes the following steps. As shown in FIG. 6, a step mirror 80 is first placed on the carrying platform 25, and then the light source module 20 is generated at least one point in step 301. A light source, which is a white light source, is projected onto the plane mirror 80 via the objective lens 21. In the embodiment of step 301, a point source of a plurality of two-dimensional arrays is illustrated. In step 31, a plurality of incident lights reflected by the light modulating unit 202 to the objective lens 21 are reflected from the plane mirror 80 to the image capturing device 22. Then, in step 302, the image capturing device 22 is controlled to capture the image, and the captured image is captured as shown in FIG. 4A. It should be noted that since the embodiment has a plurality of point light sources, the corresponding optical diffraction images are also generated with a plurality of optical diffraction patterns. Each of the captured optical diffraction patterns, as shown in Fig. 4B or Fig. 4C, differs in the length of the exposure time. 5A and FIG. 5B are computer simulation images of optical diffraction pattern light intensity corresponding to FIG. 4B and FIG. 4C, respectively.

Returning to FIG. 6, proceeding to step 303, the plane mirror 80 is subjected to depth scanning and an optical diffraction image of a corresponding depth is captured. Each depth h ₀ ~ h _n takes an optical diffraction image, and each image has a plurality of optical diffraction patterns corresponding to the position of the point source. For example, after the image capturing device 22 captures the optical diffraction image about the current plane position h ₀ , then changes the position h _{0 of} the plane mirror to h ₁ , for example, through the piezoelectric element (PZT) to the next one. After position h ₁ , continue to draw an optical diffraction image of the changed position. Due to the optical defocusing effect of the optical confocal, the optical diffraction pattern produced by each point source at each depth position is unique. After changing the depth position of the plane mirror from h ₀ to h _n , that is, when all the depths of the corrected depth range are completed, a plurality of optical diffraction patterns corresponding to different known depths can be obtained corresponding to the position of each point source. These optical diffraction patterns can be used as a comparison optical diffraction pattern sample for judging the surface depth of the object to be tested in the future. Therefore, the database stores a plurality of optical diffraction pattern samples having a position of the point light source generated by the plurality of corresponding light modulation units 202.

…..(6)It should be noted that in order to avoid the problem that the light intensity of the optical diffraction pattern obtained in step 33 is different in order to avoid the difference in the reflectance of the surface of the object or the internal light reflection interface, in another embodiment, it is further possible to A sample of the diffraction pattern for each point source is extracted and normalized. The equation is as shown in equation (6):

.....(6)

Taking the optical diffraction pattern sample shown in FIG. 7 as an example, it is an image of 5×5 pixels, wherein the oblique line area represents a bright part, and the blank area represents a dark part. Each pixel (P ₀ ~ P ₂₄ ) corresponds to a light intensity I ₀ (x, y) ~ I ₂₄ (x, y). The normalized light intensity In ₀ (x, y) ~ In ₂₄ (x, y) is calculated by equation (6). Where I _max and I _min represent the strongest light intensity values and the weakest light intensity values in I ₀ (x, y) ~ I ₂₄ (x, y) before normalization. In the example of FIG. 7, P has the strongest light intensity value of ₁₂ pixels, the weakest light intensity value of P _0. Therefore, I _max and I _{min of} equation (6) are known. Then, by substituting the light intensity of each pixel into equation (6) for calculation, the normalized light intensity value of each pixel can be obtained. Therefore, for each point source position, each depth of the optical diffraction pattern sample can be normalized, so that each pixel of the optical diffraction pattern sample has a normalized light intensity value. .

Similarly, for the combination of the dispersive objective lens and the color image capturing device, as shown in the architecture of FIG. 1B, an optical diffraction pattern sample for a plurality of wavelengths of light corresponding to different colors is established, in an embodiment. Since the color image capturing device has three color (RGB) filtering elements, it is possible to create a red optical diffraction pattern sample, a green optical diffraction pattern sample, and a blue optical diffraction pattern sample, each of which is optically wound. The shot pattern samples can all be normalized by equation (6) such that each pixel of the optical diffraction pattern sample has a normalized light intensity value. It should be noted that the establishment of the database only needs to be carried out at the beginning of the erection measurement system, and the obtained deep optical diffraction pattern sample can be used as a basis for later determining the surface depth of the object to be tested.

After the step 30, the step 31 of measuring the three-dimensional shape of the surface of the object to be tested can be performed. In this step, the control of the plurality of point source arrays and the detection rate of the image capturing device can be completed. A (quasi-single shot) full-field surface three-dimensional topography scan is performed to obtain an optically diffracted image of at least one location of the surface of the object to be tested. For example, when the objective lens 21 projects a plurality of point light sources generated by the light modulating unit 202 onto the surface of the object 8 to be tested, the image capturing device 22 can capture a diffraction image. Each of the diffracted images has a plurality of optical diffraction patterns corresponding to a plurality of point light sources. Each of the optical diffraction patterns is composed of a plurality of pixels, each of which has a corresponding light intensity value, which is obtained by the image capturing device 22.

….. (7) 如圖8所示，其中標號50代表影像擷取裝置所產生的光學繞射影像中，關於待測物表面上的特定位置被一點光源投射所反射的光學繞射圖案，而標號511至511n則代表資料庫中對應該點光源位置的多張光學繞射圖案樣本，每一光學繞射圖案樣本對應一個深度值。光學繞射影像的每一像素的亮度值I(x,y)，和每一個光學繞射圖案樣本對應的像素所具有的正規化亮度值R(x,y)，可透過方程式(7)進行演算，而得到一個NCC值。 因此，光學繞射圖案50在和複數張光學繞射樣本511~511n進行運算之後，可以得到複數個NCC0~NCCn值。由於每一張光學繞射樣本511至511n對應一個深度，因此每一個NCC 值同樣對應一個深度，藉由複數個NCC值NCC0~NCCn與其對應的深度，可以建構出如圖9所示的正規化互相關值與深度關係曲線。從曲線中可以看到其具有一最大值，該最大的NCC值代表待測物的光學繞射圖案和對應該NCC值的光學繞射圖案樣本兩者最接近。因此，最後一個步驟 33，即為從該複數個NCC值NCC0~NCCn中，找出最大NCC值，並以該NCC值所對應的深度值，作為待測物上對應該光學繞射圖案位置的深度。以圖9為例，在深度 250 μm的地方其NCC值最大，因此可以代表待測物表面上對應該光學繞射圖案50的位置其深度為250 μm。同理，其他點光源所對應的光學繞射圖案也是根據前述的方式找出相應的深度。最後，根據多個對應多個點光源位置的深度，即可以透過單一次的光學繞射影像擷取，進行物體表面或內部光反射介面全域式的形貌量測掃描，進而完整的重建待測物表面的二維或者是三維形貌。After obtaining the optical diffraction image of the object to be tested, then performing step 32 through the image comparison step, such as normalized cross correlation, to determine the depth of the corresponding point source on the object to be tested. In this step, a plurality of optical diffraction pattern samples corresponding to the position of the point source corresponding to the optical diffraction image are first selected from the database, and each of the optical diffraction pattern samples corresponds to a depth value. Next, according to equation (7), each optical diffraction pattern is extracted from the optical diffraction image, and a normalized cross-correlation value of each optical diffraction pattern and each optical diffraction pattern sample is calculated (normalized Cross correlation, NCC).

..... (7) As shown in FIG. 8, wherein reference numeral 50 represents an optical diffraction pattern reflected by a point source on a specific position on the surface of the object to be tested, in the optical diffraction image produced by the image capturing device, Reference numerals 511 to 511n represent a plurality of optical diffraction pattern samples corresponding to the position of the point source in the database, and each optical diffraction pattern sample corresponds to a depth value. The luminance value I(x, y) of each pixel of the optical diffraction image and the normalized luminance value R(x, y) of the pixel corresponding to each optical diffraction pattern sample can be performed by equation (7). Calculus, and get an NCC value. Therefore, after the optical diffraction pattern 50 is operated with the plurality of optical diffraction samples 511 to 511n, a plurality of NCC0 to NCCn values can be obtained. Since each of the optical diffraction samples 511 to 511n corresponds to a depth, each NCC value also corresponds to a depth, and the normalization shown in FIG. 9 can be constructed by a plurality of NCC values NCC0 to NCCn and their corresponding depths. Cross-correlation value versus depth curve. It can be seen from the curve that it has a maximum value which represents the closest optical diffraction pattern of the object to be tested and the optical diffraction pattern sample corresponding to the NCC value. Therefore, in the last step 33, the maximum NCC value is found from the plurality of NCC values NCC0 to NCCn, and the depth value corresponding to the NCC value is used as the position corresponding to the optical diffraction pattern on the object to be tested. depth. Taking Fig. 9 as an example, the NCC value is the largest at a depth of 250 μm, so that it can represent the position of the object to be tested corresponding to the optical diffraction pattern 50 to a depth of 250 μm. Similarly, the optical diffraction pattern corresponding to other point sources also finds the corresponding depth according to the foregoing manner. Finally, according to the depths corresponding to the positions of the plurality of point light sources, the single-mode optical diffraction image capture can be performed, and the surface topographic measurement of the surface of the object or the internal light reflection interface can be performed, and then the complete reconstruction is performed. The two-dimensional or three-dimensional shape of the surface of the object.

It should be noted that, for the architecture shown in FIG. 1B or FIG. 1C, the color image capturing device 22a can generate an optical diffraction image corresponding to a plurality of optical wavelengths, and the foregoing optical wavelength can be a single wavelength. Or it is composed of a continuous wavelength segment. In one embodiment, there are red optical diffracted images, green optical diffracted images, and blue optical diffracted images. The optical diffraction pattern for each position in the optical diffraction image of each color is calculated in the optical diffraction pattern sample corresponding to the corresponding color, and the normalized cross-correlation value and depth relationship curve corresponding to the different colors can also be obtained. Finally, the maximum NCC value is found from the plurality of curves, and the depth value corresponding to the NCC value is used as the depth corresponding to the position of the optical diffraction pattern on the object to be tested.

In addition, it should be noted that if the measurement resolution is to be increased, the position of the point source may be changed again, as shown in FIG. 10, wherein at the first time point, the position of the point source modulated by the light modulation unit 202 is As shown in the left area, in the state where 203a is turned on, the light is guided to the objective lens, and when the state of 203b is off, the light is guided to other places. When the optical diffraction image is captured, the control and calculation unit controls the light modulation unit 202 to change the position of the point light source to form a state as described in the right region of FIG. 10, so that different positions of the object table can be measured. Perform depth measurements to increase resolution.

In summary, since the optical system of the present invention and the surface of the object or the internal light reflection interface (including the back side of the object) topography detection method can utilize the digital light source to perform synchronous multi-point optical scanning for high-speed global surface three-dimensional surface detection, And solving the problem of optical crosstalk caused by the three-dimensional shape of the surface of the confocal detection object or the internal light reflection interface (including the back side of the object), thereby improving the detection accuracy and efficiency, and also avoiding the need for the horizontal or The effect of the vibration generated during vertical scanning on the detection effect.

The above description is only intended to describe the preferred embodiments or embodiments of the present invention, which are not intended to limit the scope of the invention. That is, the equivalent changes and modifications made in accordance with the scope of the patent application of the present invention or the scope of the invention are covered by the scope of the invention.

2‧‧‧Optical system
20‧‧‧Light source module
200‧‧‧Light source
201‧‧‧ collimating mirror
202‧‧‧Light Modulation Unit
203, 203a, 203b‧‧‧ light switch
21, 21a‧‧‧ objective lens
22, 22a‧‧‧ image capture device
23‧‧‧Control and calculation unit
24‧‧‧Spectral components
25‧‧‧Loading station
26‧‧‧ Focusing mirror
50‧‧‧ optical diffraction pattern
511~511n‧‧‧ optical diffraction pattern sample
8, 8a‧‧‧Test objects
80‧‧‧ flat mirror
81‧‧‧ bottom
82‧‧‧Internal plane
83‧‧‧ Cracks inside the object to be tested or hollow structures inside
90‧‧‧ incident light
91‧‧‧ Collimated light field
93, 93a, 93b‧‧‧ ‧ light

1A is a schematic view showing a first embodiment of an optical system according to the present invention; FIG. 1B is a schematic view showing a second embodiment of an optical system according to the present invention; FIG. 1C is a schematic view showing a third embodiment of the optical system of the present invention; FIG. 2A and FIG. 2B are diagrams showing the light control switch and changing its reflection direction; FIG. 2C is a schematic diagram of simulating a point light source by using a light modulation unit; FIG. 3 is a schematic diagram showing the diffraction of Fulang and Fei; FIG. The diffracting device captures a diffracted image of an object surface or an internal light reflecting interface; FIGS. 4B and 4C are optical diffraction images of different exposure times; FIGS. 5A and 5B are optical windings of FIGS. 4A and 4B 3D computer simulation diagram of image light intensity; Figure 6 is a schematic diagram of the construction of the database and the position of the plane mirror; Figure 7 is a schematic diagram of the light intensity of each pixel in the optical diffraction pattern sample; Figure 8 is a diagram of each optical diffraction pattern and corresponding Schematic diagram of the calculation relationship of a plurality of optical diffraction pattern samples; FIG. 9 is a schematic diagram of normalized cross-correlation and depth relationship curves; and FIG. 10 is a light modulation unit transformation point Schematic diagram of the source.

2‧‧‧Optical system

20‧‧‧Light source module

200‧‧‧Light source

201‧‧‧ collimating mirror

202‧‧‧Light Modulation Unit

203‧‧‧Light switch

21‧‧‧ Objective lens

22‧‧‧Image capture device

23‧‧‧Control and calculation unit

24‧‧‧Spectral components

25‧‧‧Loading station

8‧‧‧Test object

90‧‧‧ incident light

91‧‧‧ Collimated light field

93‧‧‧ ‧ light

Claims (24)

An optical system includes: a light source module for generating at least one light source; an objective lens for projecting the at least one light source to at least one position of an object surface or an internal light reflecting interface; an image capturing device, Receiving at least one reflector light reflected by the at least one location to generate a corresponding optical diffraction image having at least one optical diffraction pattern, each optical diffraction pattern corresponding to an object surface or internal light reflection a topographical depth position on the interface; and a control and calculation unit for calculating an optical diffraction pattern corresponding to each depth position of the optical diffraction image and a plurality of optical diffraction pattern samples to determine each The depth of the topography of the surface of the object or the position of the internal light reflecting interface. The optical system of claim 1, wherein the light source module further comprises: a light source for generating an incident light; a collimating mirror group for collimating the incident light; and a light modulation And a unit for modulating the collimated incident light into the at least one light source. The optical system of claim 2, further comprising a beam splitting element disposed between the collimating mirror group and the light modulating unit, wherein the objective lens is a dispersive objective lens, wherein the incident light passes through the optical lens The light splitting component is projected to the light modulating unit, and the light modulating unit modulates the incident light into at least one light source, and then reflects the light to the collimating mirror group, and then guides the at least one light source to the color objective lens, and the dispersion After the objective lens disperses the point source and then projects onto the surface of the object or the internal light reflection interface, the reflector light is generated, and the reflector light is again projected to the light modulation unit according to the original light path, and the light modulation unit further reflects the reflector Light is reflected to the spectroscopic element, and the spectroscopic element directs the spectroscopic light to the image capturing device. The optical system of claim 2, wherein the light modulating unit is a digital micro mirror element or a germanium based liquid crystal element. The optical system of claim 2, wherein the incident light is broadband light or single frequency light. The optical system of claim 1, wherein the control and calculation unit operates each optical diffraction pattern and the plurality of optical diffraction pattern samples such that each optical diffraction pattern is generated relative to the plurality of optical diffraction patterns. a plurality of normalized cross-correlation values of the optical diffraction pattern samples, each of the normalized cross-correlation values corresponding to a depth, and the control and the calculation unit have corresponding depths corresponding to the maximum normalized cross-correlation values, corresponding to optical diffraction The depth of the topography of the surface of the patterned object or the position of the internal light reflecting interface. The optical system of claim 1, wherein the at least one light source constitutes a one-dimensional or two-dimensional array of point light sources. The optical system of claim 1, wherein each point source is at a distance from each other without crosstalk. The optical system of claim 8, wherein the control and calculation unit controls the light source module to generate a plurality of point light sources to complete a three-dimensional topography scan of the quasi-click surface, thereby obtaining a An optically diffracted image of a plurality of locations on the surface of the object. The optical system of claim 1, wherein the objective lens is a chromatic objective lens, and the image capturing device is a color image capturing device for generating a plurality of optical diffraction images, each of the optical diffraction images. An optical diffraction image formed by a light of a wavelength corresponding to a depth focus position thereof. The optical system of claim 10, wherein the dispersive objective lens is a refractive or optically diffractive dispersive objective lens. A method for detecting a three-dimensional shape of an object surface or an internal light reflection interface, comprising the steps of: generating at least one light source; projecting the at least one light source to an at least one topography depth position of an object surface or an internal light reflection interface by an objective lens Passing at least one reflector light reflected by the at least one topography depth position through the objective lens; guiding the at least one reflector light to an image capturing device; the image capturing device sensing the at least one reflector light Generating an optical diffraction image having at least one optical diffraction pattern corresponding to a depth position of at least one topography; and extracting at least one optical diffraction pattern from the optical diffraction image, and each optical The diffraction pattern is computed with a plurality of optical diffraction pattern samples to determine the depth of the topography corresponding to the position of each object surface or internal light reflecting interface. The method for detecting a three-dimensional shape of an object surface or an internal light reflection interface according to claim 12, wherein the at least one light source constitutes a one-dimensional or two-dimensional array of point light sources. The method for detecting a three-dimensional shape of an object surface or an internal light reflection interface according to claim 12, wherein the generating the at least one light source further comprises the steps of: providing a light source for generating an incident light; The mirror group collimates the incident light and directs the collimated incident light to a light modulation unit; and controls the first region of the light modulation unit to modulate the collimated broadband light into the at least one light source, and guides To the objective lens, and controlling a second region of the light modulation unit that surrounds the periphery of the first region, directing the collimated incident light to a region different from the objective lens. The method for detecting a three-dimensional shape of an object surface or an internal light reflection interface according to claim 12, wherein the generating the at least one light source further comprises the steps of: providing a light source for generating an incident light; The mirror group collimates the incident light and directs the collimated incident light to a light modulation unit; and controls the light modulation unit to generate a plurality of point light sources to complete the quasi-click global surface three-dimensional topography scan. The method for detecting a three-dimensional shape of an object surface or an internal light reflection interface according to claim 12, wherein the generating the at least one light source further comprises the steps of: providing a light source for generating an incident light; The mirror group collimates the incident light and directs the collimated incident light to a beam splitting element; the beam splitting element splits and directs the incident light to a light modulating unit; and controls the first region of the light modulating unit to And collimating the broadband light into the at least one light source, guiding to the dispersion objective lens, and controlling a second region of the light modulation unit surrounding the periphery of the first region, guiding the collimated incident light to be different from the The area of the dispersion objective. The method for detecting a three-dimensional shape of an object surface or an internal light reflection interface according to claim 12, wherein the generating the at least one light source further comprises the steps of: providing a light source for generating an incident light; The mirror group collimates the incident light and directs the collimated incident light to a light splitting element; the splitting element splits and directs the incident light to a light modulating unit; and controls the light modulating unit to generate a plurality of point light sources Complete the three-dimensional topography scan of the surface of the universal object or the internal light reflection interface of the quasi-click. The method for detecting a three-dimensional shape of an object surface or an internal light reflection interface according to claim 16 further comprising the step of modulating the light of the reflector: controlling the first region of the light modulation unit to reflect the reflector Light is guided to the light splitting element, and controls a second region of the light modulating unit that surrounds a periphery of the first region, directs the reflector light to a region different from the light splitting element; and the light splitting element reflects the light The object light is split and directed to the image capture device. The method for detecting a three-dimensional shape of an object surface or an internal light reflection interface according to claim 15, wherein the light modulation unit is a digital micro mirror element or a germanium liquid crystal element. The method for detecting a three-dimensional shape of an object surface or an internal light reflection interface according to claim 15 , wherein the incident light is broadband light or single frequency light. The method for detecting a three-dimensional shape of an object surface or an internal light reflection interface according to claim 12, wherein determining the depth of the surface of each object or the position of the internal light reflection interface further comprises the following steps: a library having sample information of at least one wavelength of light, each sample information comprising a plurality of optical diffraction pattern samples corresponding to different depths; each of the extracted optical diffraction patterns corresponding to the data library a plurality of optical diffraction pattern samples corresponding to different depths of the point source position are calculated to generate a plurality of cross-correlation values for the optical diffraction pattern samples; and optical windings corresponding to the maximum of the plurality of cross-correlation values The depth corresponding to the pattern sample is taken as the topography depth corresponding to the position of the optical diffraction pattern. The method for detecting a three-dimensional shape of an object surface or an internal light reflection interface according to claim 20, wherein the establishing the sample information of each light wavelength further comprises the following steps: providing a mirror surface having a surface and the objective lens a first depth distance; at least one correction point light source corresponding to the specific light wavelength is generated by the object lens and then incident on the mirror surface; and the reflector light reflected from the specular lens is guided to the image through the objective lens. Taking a device to generate a corrected optical diffraction image about the first depth distance, which has at least one optical diffraction pattern sample respectively corresponding to a correction point light source; respectively extracting at least the first depth distance An optical diffraction pattern sample; changing the first distance to a second depth distance to further obtain another corrected optical diffraction image about the changed second depth distance; and separately extracting the second depth distance At least one optical diffraction pattern sample. The method for detecting a three-dimensional shape of an object surface or an internal light reflection interface according to claim 21, further comprising the step of performing normalization calculation on each optical diffraction pattern sample separately. The method for detecting a three-dimensional shape of an object surface or an internal light reflecting interface according to claim 21, wherein the objective lens is a chromatic objective lens, and the image capturing device is a color image capturing device for generating a plurality of color image capturing devices. Corrected optical diffraction images corresponding to different wavelengths of light.