TWI886056B - Line-scan chromatic confocal system having full-field deconvolution for reconstructing surface profile - Google Patents

Abstract

A line-scan chromatic confocal system with full-field deconvolution for reconstructing surface profile comprises a light source module for generating a detecting light, a first spatial filter, an image module, and a spectrum analyzing device. The first spatial filter filters the detecting light to form a filtered detecting light. The dispersing module disperses the filtered detecting light to form a dispersed light projecting onto a sample thereby forming an object light reflecting from the sample. The spectrum analyzing device receiving and analyzing the object light further comprises a second spatial filter, a spectrometer, and a processing unit. The object light is filtered by the second spatial filter so as to form a filtered object light. The spectrometer receives the filtered object light and generates a spectrum information corresponding to the filtered object light. The processing unit having a point spread function (PSF) receives the spectrum information and performs a deconvolution calculation using the PSF and spectrum information for reconstructing a surface profile image with respect to the sample.

Description

Line Scanning Color Confocal Measurement System with Global Deconvolution Surface Topography Reconstruction

The present invention is a color confocal measurement technology, in particular, a line scanning color confocal measurement system with global deconvolution surface morphology reconstruction and deconvolution calculation to improve measurement resolution.

Monochromatic and color confocal systems have undergone a fairly mature development process since 1961, covering theoretical research, engineering practice, and even commercial applications. This technology has been widely used in the fields of biomedicine and industrial testing, and has become one of the important technologies for micron-level optical morphology reconstruction.

In the development of color confocal technology, the linear color confocal system has gradually become a focus of attention. First, the linear color confocal system uses a single-axis scanning motion, without the need for complex dual-axis scanning, making its scanning speed faster. This not only enables the rapid acquisition of a large range of confocal images, but also enables efficient measurement in a short time. Especially for industrial inspection applications, the advantage of linear scanning is that it saves time and improves production efficiency. Secondly, since the linear color confocal only requires single-axis scanning, compared to the three-axis scanning required by the single-point monochrome confocal system, its scanning dimension requirements are lower. The optical configuration and mechanical design are also simpler, and these advantages together support the prospect of linear chromatic confocal microscopy becoming the future non-contact online measurement method at the micron and sub-micron level.

Despite the aforementioned advantages, linear chromatic confocal systems still have problems to be solved, such as relatively poor resolution in the line direction, which is usually called lateral crosstalk. This may cause some blur or distortion in the measurement of sample surface morphology, especially in the measurement of microstructures. This is a problem that cannot be ignored for applications such as high-precision industrial inspection.

The original confocal system has excellent optical depth sectioning and lateral resolution capabilities, but its measurement speed is relatively slow due to the need for three-axis scanning. To overcome this shortcoming, researchers have proposed different optical architectures: using lens arrays, pinhole arrays or slits to replace traditional pinholes to achieve lateral parallel measurement; using galvanometers or spatial light modulators to replace traditional displacement platforms to increase the lateral scanning speed of the confocal system; developing color confocal technology to create axial dispersion and eliminate the need for axial scanning. Although these new architectures have their own limitations, these studies have promoted the development of confocal microscopy, expanding its application areas to high-speed biomedical detection and microstructure measurement, showing great potential.

In order to overcome the lateral interference problem of the aforementioned linear color confocal system, the present invention provides a line scanning color confocal measurement system with global deconvolution surface morphology reconstruction, which uses a spatial light modulator (SLM) such as Liquid Crystal On Silicon (LCoS) or a combination of filter elements of various sizes to replace the detection end slit to most directly measure the point spread function of the spectral end with spatial variation. By measuring the spatially variable point spread function, combined with the calculation of the basis deconvolution, the surface image of the object to be measured can be effectively restored to solve the problem of lateral crosstalk.

In one embodiment, the present invention provides a line scanning color confocal measurement system with full-domain back-convolution surface morphology reconstruction, including a light source module that generates linear detection light, a first spatial filter, a dispersion module, and a spectrum analysis device. The first spatial filter filters the linear detection light into linear filtered light. The dispersion module disperses the linear filtered light to form dispersed light that is projected onto the object to be measured and then reflected to form measurement light. The spectrum analysis device is used to receive and analyze the measurement light, and the spectrum analysis device further has a second spatial filter, a spectrometer, and an operation processing unit. The second spatial filter filters the measurement light to form filtered measurement light. The spectrometer receives filtered object measurement light to generate corresponding spectral information. The computational processing unit has a point spread function, receives the spectral information, and performs a deconvolution operation on the point spread function and the spectral information to reconstruct the surface morphology image of the object to be measured.

2: Line scanning color confocal measurement system

20: Light source module

200: Light source

201: Collimator

202: Modulation lens set

202a: Beam expander set

202b: Achromatic lens set

202c, 202d: laminated convex lens

220: Collimator lens set

221:Spectroscopic element

222: Vibration mirror element

223: First Scanning Mirror Set

224:Dispersive objective lens

225: Second scanning lens set

226: Linear polarizer

21: First Space Filter

22:Dispersion module

23: Spectral analysis device

24: Cylindrical concave lens

25: Diffusion film

230: Polarization beam splitter element

231, 231a: Second space filter

2310: Filter plate

2311: Position adjustment element

2312, 2313: Filter structure

2312a: First pinhole array

2312b: Second pinhole array

2313a: The first narrow seam

2313b: The second narrow seam

232: Spectrometer

233: Operation processing unit

90:Emits a beam of light

91: Linear detection light

92: Linear filter light

93: Measuring object light

94: Filtering the object measurement light

92a: Linear scattered light

920~922: Linear scattered light

S: Object to be tested

SP: Platform

FP: front focal plane

4: Methods

40~42: Steps

400~406: Steps

410~418: Steps

420~427: Steps

FIG1 is a schematic diagram of an embodiment of the line scanning color confocal measurement system with full-area backconvolution surface topography reconstruction of the present invention.

Figure 2A is a schematic diagram of optical imaging using a flat-field scanning lens.

Figure 2B is a schematic diagram of optical imaging of the F-θ scanning lens.

Figure 3A is a schematic diagram of a non-flat field non-telecentric optical structure. Figure 3B is a schematic diagram of a flat field non-telecentric optical structure.

Figure 3C is a schematic diagram of the telecentric optical structure of the dispersive objective lens of the present invention at the object end.

FIG4 is a schematic diagram of an embodiment of the second spatial filter formed by combining filter elements of various sizes according to the present invention.

Figure 5A is a schematic diagram of a process of the line scanning color confocal measurement method of the present invention.

FIG5B is a schematic diagram of a process for obtaining a response curve in the line scanning color confocal measurement method of the present invention.

FIG5C is a schematic diagram of a process for obtaining a point spread function in the line scanning color confocal measurement method of the present invention.

Figures 6A to 6D are schematic diagrams of controlling the second spatial filter to simulate different pinhole positions when performing point spread function imaging of the present invention.

FIG7 is a schematic diagram of a process of measuring an object to be measured by the line scanning color confocal measurement method with full-area backconvolution surface morphology reconstruction of the present invention.

Various exemplary embodiments will be more fully described below with reference to the accompanying drawings, some of which are shown in the accompanying drawings. However, the concepts of the present invention may be embodied in many different forms and should not be construed as limited to the exemplary embodiments described herein. Rather, these exemplary embodiments are provided so that the present invention will be detailed and complete and will fully convey the scope of the concepts of the present invention to those skilled in the art. Similar numbers always indicate similar elements. The line scan color confocal measurement system will be described below with a variety of embodiments in conjunction with the drawings, however, the following embodiments are not intended to limit the present invention.

Please refer to FIG. 1, which is a schematic diagram of an embodiment of the line scanning color confocal measurement system with global backconvolution surface topography reconstruction of the present invention. In this embodiment, the line scanning color confocal measurement system 2 includes a light source module 20, a first spatial filter 21, a dispersion module 22 and a spectrum analysis device 23. The light source module 20 includes a light source 200, a collimator 201 and a modulator group 202. In one embodiment, the light source 200 can be a broadband light source including a xenon lamp, a light-emitting diode (LED), a halogen lamp, a laser driven light source (LDLS), or a supercontinuum laser, and is located in the 450nm~650nm visible light band. A light source with a smooth and uniform spectrum distribution is a better choice, but it is not limited to this.

The light beam 90 emitted by the light source 200 is first collimated by the collimator 201, and then passes through the modulator group 202, which is used to modulate the collimated light beam into a linear detection light and modulate the divergence angle of the linear detection light in the linear length direction, and focus it to the first spatial filter 21. In this embodiment, the modulator group 202 further includes a beam expander group 202a and an achromatic lens group 202b, wherein the light beam passing through the collimator 201 first passes through the beam expander group 202a, and then passes through the achromatic lens group 202b to form a linear detection light 91, and then moves forward along the optical path in the direction of the object to be measured. In this embodiment, the beam expander set 202a is a Galilean beam expander, which is composed of two sets of double-collagen convex lenses 202c and 202d, and is used to enlarge the diameter of the light beam 90 to the size required for measurement. The achromatic lens set 202b is a cylindrical lens in this embodiment, but is not limited to this. The achromatic lens set 202b modulates the energy distribution of the point light source into a focused linear detection light 91, and then guides the focused linear detection light 91 to the first spatial filter 21. In this embodiment, the first spatial filter 21 further includes a cylindrical concave lens 24 and a diffuser 25, wherein the cylindrical concave lens 24 matches the numerical aperture required for imaging, and the diffuser 25 ensures that after passing through the diffuser 25, a plurality of point light sources with specific numerical apertures are formed on the first spatial filter 21, and are divergently projected to the dispersion module 22.

After the linear detection light 91 passes through the first spatial filter 21, it becomes a linear filtered light 92. In the present embodiment, the first spatial filter 21 is a slit structure, but it is not limited to this. Its structure and shape are determined according to the measurement requirements. The linear filtered light 92 is projected to the object to be measured through the dispersion module 22, and then reflected to form the object measurement light 93, and then guided to the spectrum analysis device 23 through the dispersion module 22 along the original light path. In the present embodiment, the dispersion module 22 includes a collimator lens group 220, a spectrometer element 221, a galvanometer element 222, a first scan lens group (scan lens) 223 and a dispersion objective lens 224 in sequence. In this embodiment, the linear filtered light 92 is collimated by the collimating lens set 220 and then split by the beam splitter 221. In this embodiment, the beam splitter 221 is a non-polarized beam splitter (NPBS).

The split linear detection light is then reflected by the galvanometer element 222 to the first scanning lens group 223. In this embodiment, the galvanometer element 222 is used to adjust the reflection angle of the linear detection light by rotating to change the angle θ, so as to change the position of the projection to the object S to be detected. In this embodiment, the galvanometer element 222 is placed at the entrance pupil of the first scanning lens group 223 to receive the linear detection light. The size of the mirror surface of the galvanometer element 222 must match the entrance pupil of the first scanning lens group 223. Since the present embodiment uses a galvanometer element 222, a first scanning lens group 223 is added between the galvanometer element 222 and the dispersive objective lens 224 to focus the linear filtered light 92 at the focal plane FP before the dispersive objective lens 224, and then enter the dispersive objective lens 224. The first scanning lens group 223 can be regarded as a relay lens, which increases the concentric focus, so that the system is not limited by space, and it is convenient to perform beam scanning.

In this embodiment, the first scanning lens group 223 is an F-θ scanning lens (F-θ Scan Lens) that has both chromatic aberration elimination and telecentricity characteristics. As shown in FIG2A and FIG2B, FIG2A is a schematic diagram of optical imaging of a conventional flat-field scanning lens; FIG2B is a schematic diagram of optical imaging of an F-θ scanning lens. Compared with a generally distortion-free flat-field scanning lens 100, the F-θ scanning lens 223 is designed so that the lens group has a specific distortion (Distortion), so that the image height (h) changes from h=F * tan θ to a linear relationship between the image height and the incident angle h=F *θ, where h is the image height, F is the focal length of the scanning lens, and θ is the scanning angle. Since the conventional flat-field scanning lens 100 has a flat-field focal plane, the scanning angle and the imaging position are not linear, but a tangent relationship, h=F*tanθ. When using a general flat-field scanning lens, the image height changes nonlinearly (F*tanθ), and the uniform vibration mirror movement will cause the exposure time to change with the scanning position (image height). Therefore, the problem of the exposure time changing with the scanning position (image height) when using a conventional flat-field scanning lens can be solved by using an F-θ scanning lens.

Returning to FIG. 1 , the linear detection light passing through the first scanning lens group 223 and then passing through the dispersive objective lens 224 makes the linear filtered light 92 dispersed to form a linear dispersed light 92a, which has a plurality of linear dispersed lights 920-922 with different focal depths projected onto an object to be measured S, and reflected from the object to be measured S to form a linear object measurement light 93. In FIG. 1A , the RGB three-color light 920-922 is represented. In this embodiment, the dispersive objective lens 224 is a dispersive objective lens with a double telecentric structure. In optics, telecentricity is defined as the entrance pupil or exit pupil of a system being at infinite distance. If the entrance pupil is at infinite distance, the system is object-side telecentric; if the exit pupil is at infinite distance, it is image-side telecentric; if both are at infinite distance, it is a double-telecentric system, that is, the principal rays of the incident and outgoing light are parallel to the optical axis. The characteristics of object-side telecentricity are shown in Figures 3A to 3C, where Figure 3A is a schematic diagram of a non-flat non-telecentric optical structure; Figure 3B is a schematic diagram of a flat non-telecentric optical structure; and Figure 3C is a schematic diagram of the object-side telecentric optical structure of the dispersive objective lens of the present invention. In FIG3A , the linear detection light entering the dispersive objective lens 224 from the scanning lens produces a field curvature effect, so that each color light of the same wavelength on the linear detection light has a different focus depth, as shown by curve 99 in FIG3A . Although the dispersive objective lens 224 in FIG3B has a flat field effect, the off-axis light 92' on both sides has an incident angle relative to the surface of the object to be measured S, so that the reflected object light 93 cannot enter the dispersive objective lens 224, resulting in the loss of measurement light information. As shown in FIG3C , the dispersive objective lens 224 of the present invention makes the magnification of all depths equal, and the field curvature phenomenon of the off-axis point is smaller, thereby reducing the measurement error of the full-field system. In addition, since each of the linear dispersion lights 920-922 has a plurality of sub-dispersion lights, for example, the linear dispersion light 920 is composed of sub-dispersion lights 9200, 9201, 9202, the linear dispersion light 921 is composed of sub-dispersion lights 9210, 9211, 9212, and the linear dispersion light 922 is composed of sub-dispersion lights 9220, 9221, 9222. Each sub-dispersion light has an optical axis SOA. The dispersive objective lens of the present invention has the characteristics of being telecentric at the object end and the image end. Therefore, for each sub-dispersed light 9200~9202, 9210~9212, 9220~9222, its optical axis SOA is parallel to the central optical axis OA of the dispersive objective lens 224, so that the object light reflected from the object to be measured S can be completely received by the dispersive objective lens and returned to the optical system, thereby increasing the optical efficiency of the system.

Returning to FIG. 1 , the object light 93 reflected from the object to be measured passes through the dispersion objective lens 224, the first scanning lens group 223, the oscillating lens element 222, the beam splitter element 221, the second scanning lens group 225, and the linear polarizer 226 in sequence. After the object light 93 passes through the linear polarizer 226, almost only the S-polarized object light passes through. Afterwards, the S-polarized object light enters the spectrum analysis device 23. In this embodiment, the spectrum analysis device 23 further includes a polarization beam splitter element 230, a second spatial filter 231, a spectrometer 232, and an operation processing unit 233. The polarization splitting element 230 polarizes and splits the object light, and reflects it to the second spatial filter 231, so as to filter the object light to form filtered object light 94. In this embodiment, the second spatial filter 231 is Liquid Crystal On Silicon (LCoS).

Since the S-polarized object light is perpendicular to the beam splitting interface of the polarization beam splitter 230, and the P-polarized object light is parallel to the beam splitting interface, most of the S-polarized object light is then reflected by the polarization beam splitter 230 to the second spatial filter 231. When the pixel unit of the panel of the second spatial filter 231 is in the closed state, the polarization state of the light reflected from the panel of the second spatial filter 231 remains unchanged and is still S-polarized. When the object light is detected by the S-polarization state and enters the polarization spectrometer 230 again, most of it is reflected by the spectroscopic interface and cannot penetrate the polarization spectrometer 230. On the contrary, when the pixel unit is in the on state, the polarization state of the reflected light changes due to the liquid crystal arrangement, so that it changes from the S-polarization state object light to the P-polarization state object light. Similarly, after passing through the spectroscopic interface of the polarization spectrometer 230, most of the light will penetrate and enter the subsequent spectrometer 232. The spectrometer 232 receives the filtered object light to generate corresponding spectral information. In this embodiment, the spectrometer 232 includes a light-guiding element 2320 to reflect the object light to the image sensing element 2321, generate a corresponding spectral image and transmit it to the calculation processing unit 233. The calculation processing unit 233 has a point spread function. The calculation processing unit 233 receives the spectral information and performs a deconvolution operation on the point spread function and the spectral information to reconstruct the surface morphology image of the object to be measured.

It should be noted that, in addition to using LCoS as mentioned above, the second spatial filter 231 can also be as shown in FIG4, which is composed of a combination of filter elements of various sizes to replace the detection end slit to most directly measure the point spread function of the spectrum end with spatial variation. By combining the measured spatial variation point spread function with the calculation of deconvolution, the surface image of the object to be tested can be effectively restored to solve the problem of lateral crosstalk. In FIG4, the second spatial filter 231a includes a filter plate 2310 and a position adjustment element 2311. The filter plate 2310 has a plurality of filter structures 2312 and 2313 of different sizes or shapes, such as a plurality of slits or a plurality of pinhole arrays of different diameters.

In one embodiment, the slits include a first slit 2313a and a second slit 2313b, and the pinhole array includes a first pinhole array 2312a with a first aperture size and a second pinhole array 2312b with a second aperture size, wherein the size of the first slit 2313a corresponds to the first pinhole array 2312a, for example, the slit width is equal to the pinhole diameter, and the size of the second slit 2313b corresponds to the second pinhole array 2312b, for example, the slit width is equal to the pinhole diameter. In this embodiment, the first pinhole array 2312a has three rows, and each pinhole corresponds to a sensing unit, for example: a sensing unit composed of at least one pixel, and the distance between rows is one sensing pixel; similarly, the second pinhole array 2312b has two rows, and its pinhole size is larger than the pinhole size of the first pinhole array 2312a, and the distance between rows is one sensing pixel. The first and second pinhole arrays 2312a and 2312b are used to obtain the point spread function of each measurement position when the line detection light is projected onto the object to be measured. The more rows, the smaller the pinhole aperture, and the greater the resolution of measuring the surface of the object to be measured. The number of rows of the pinhole array in Figure 4 is only an embodiment and is not limited to the number of rows.

The first and second slits 2313a and 2313b are slits used for actually measuring the object to be measured. For example, the point spread function obtained using the first pinhole array 2312a is applicable to the first slit 2313a for measurement, and the point spread function obtained using the second pinhole array 2312b is applicable to the second slit 2313b for measurement. The position adjustment element 2311 is connected to the filter plate 2310, and the position adjustment element 2311 is used to adjust the position of the filter plate 2310 so that one of the filter structures on the filter plate 2310 receives the object light. In this embodiment, the position adjustment element 2311 is a linear slide rail.

Since the conversion rate of LCoS in polarization state is not 100%, additional background light will be introduced in the simple shooting process. To solve this problem, in one embodiment, the LCoS is turned off for background shooting. Two shots are taken in the same state (i.e., the rotation angle and scanning depth of the galvanometer element are fixed). The first shot is to obtain the spectral image with the LCoS turned on, and the second shot is to take the photo with the LCoS turned off. Finally, the information of the two images is subtracted to obtain the final spectral image without background light. This step ensures that the spectral image obtained is not affected by the background light when measuring or calibrating.

The process of performing the line scan color confocal measurement method using the line scan color confocal measurement system of FIG1. As shown in FIG5A, the figure is a schematic diagram of a process of the line scan color confocal measurement method of the present invention. In method 4 of the present embodiment, using the system shown in FIG1, step 40 is first performed to establish a global correction spectrum. It should be noted that because the distance, tilt, and centering of optical elements in a real system are not perfect, aberrations such as telecentricity and field curvature are inevitable, so the optical performance of the measured visible range (field of view, FOV) is not consistent, and global correction for different measurement points is an inevitable means. Therefore, through step 40, using the mirror as the object to be measured, a global correction spectrum at different depths can be established.

In an embodiment of step 40, the process is shown in FIG5B. In FIG5B, firstly, step 400 is performed to adjust the second spatial filter 231 into a slit structure. In this step, because the second spatial filter adopts LCoS, the second spatial filter 231 can be controlled by signal to simulate a slit structure, that is, the pixel unit in the slit area is open, and the pixel unit in the rest of the area is closed. Then, step 401 is performed to move the carrying platform SP carrying the object to be tested to a depth. Then, step 402 is performed to scan the object to be tested, and multiple scan images corresponding to different scanning positions at corresponding depths are obtained, and each scan image has spectral information corresponding to the scanning position. In step 402, the object to be tested is a mirror structure, and the method of controlling the scanning position is to change the angle of the galvanometer element 222 so that the linear dispersion light 92a scans the object to be tested, and a scan image can be obtained at each scanning position. Therefore, after performing a plane scan of the object to be tested S, multiple scan images can be obtained. Afterwards, step 403 is performed to obtain multiple background spectral images corresponding to different scanning positions at the depth. This is because the effect of LCoS using polarization state splitting is limited, and the stray light and background light introduced are recorded. Therefore, the signal when the slit is closed needs to be subtracted from the original image to obtain an image with a theoretical background grayscale value of 0. In step 403, the second spatial filter 231 is first controlled to close the slit, and then the angle of the galvanometer element 222 is changed so that the linear dispersion light 92a scans the object to be tested, so as to obtain multiple background spectral images at different scanning positions.

Then, step 404 is performed to obtain the background-removed spectral images at different scanning positions at the depth. In step 404, the spectral information of the multiple scanned images corresponding to different scanning positions in step 402 and the background spectral information of the multiple background spectral images corresponding to different scanning positions in step 403 are subtracted to obtain multiple background-removed spectral images corresponding to different scanning positions. For example, in one embodiment, step 402 obtains three scanned images corresponding to positions A to C, and step 403 also obtains three background spectrum images corresponding to positions A to C. Then, in step 404, the scanned image corresponding to position A is subtracted from the background spectrum image to obtain the background-removed spectrum image corresponding to position A. Similarly, positions B and C are processed in the same way, so three background-removed spectrum images corresponding to positions A to C can be obtained in step 404. Then, step 405 is performed to confirm whether the depth scan is completed. If not, step 406 is performed to change to the next depth, and then steps 401 to 405 are repeated until all depth scans are completed. It should be noted that the system's calibration process uses the mirror as the object to be tested. Under the process of Figure 5B, a full-domain calibration spectrum image collection at different depths can be obtained.

Returning to step 5A, after step 40, step 41 is performed to establish a point spread function for each sensing position on the second filter. The point spread function acquisition process is performed by simulating the role of a pinhole array through the second spatial filter 231 (LCoS in this embodiment) and using a mirror as the object to be tested. In step 41, as shown in FIG5C, the figure is a schematic diagram of an embodiment of the proposed point spread function of the present invention. In this embodiment, step 410 is first performed to make the second spatial filter 231 virtualize a pinhole array. In the embodiment of step 410, as shown in FIG6A, the second spatial filter 231 simulates a one-dimensional pinhole array according to the control signal. In this embodiment, there are 16 pinholes, but this is not a limitation. Then, step 411 is performed to control the second spatial filter 231 to open multiple pinholes from the pinhole array, wherein there is at least one closed pinhole between any two opened pinholes. In step 411, as shown in FIG6B, the pinholes at positions 1, 4, 7, 10, 13, and 16 are opened first.

Then, step 412 is performed to move the mirror to the initial depth, and then step 413 is performed to capture the test spectrum image at the depth, wherein each test spectrum image is obtained by subtracting the image of the pinhole array corresponding to step 411 and the background spectrum image obtained by closing the pinhole in step 411. In this step, (1) the linear dispersion light 92a in the optical system of FIG1 is projected onto the mirror to be tested, and the formed test light enters the second spatial filter 231 shown in FIG6B, and is then imaged by the spectrometer 23. After that, (2) the pinhole in FIG6B is closed to obtain the corresponding background spectrum image. In another embodiment, the above-mentioned procedures (1) and (2) can be performed multiple times, and then the spectral information of the images obtained in each (1) and (2) is subtracted to form multiple sets of test spectral images, and finally the multiple sets of test spectral images are averaged as the test spectral images of step 413. After step 413, step 414 is performed to confirm whether the test spectral images of all pinholes are completed. If not, step 415 is performed to change the position of the pinhole, and then the above-mentioned steps 413~414 are repeated. In this step, as shown in Figures 6C and 6D, by changing the position of the pinhole, the test spectral images of all pinholes with respect to the point spread function can be obtained.

It should be noted that the aforementioned step 413 is an example of using LCoS as the second spatial filter. In another embodiment, if the method shown in FIG. 4 is used, when performing step 413, taking the first pinhole array 2312a of FIG. 4 as an example, the first row R1 of the first pinhole array 2312a is first moved to the sensing position, and then (1) the linear dispersion light 92a in the optical system of FIG. 1 is projected onto the mirror surface to be tested. The formed test light enters the first row R1 of the first pinhole array 2312a of the second spatial filter 231a shown in FIG. 4, and is then imaged by the spectrometer 23. After that, (2) closing the pinholes in FIG. 4 is performed, for example, the area without pinholes on the second spatial filter 231a in FIG. 4 is moved to the sensing position to obtain the corresponding background spectrum image. In another embodiment, the above-mentioned procedures (1) and (2) can be performed multiple times, and then the spectral information of the images obtained by (1) and (2) are subtracted each time to form multiple groups of test spectrum images, and finally the multiple groups of test spectrum images are averaged as the test spectrum image of step 413. After step 413, step 414 is performed to confirm whether the test spectrum images of all pinholes are completed. If not, then proceed to step 415 to change the position of the pinhole, that is, move the second row R2 of the first pinhole array 2312a to the sensing position, and then repeat the aforementioned steps 413~414. Then, proceed to step 415 to change the position of the pinhole, that is, move the third row R3 of the first pinhole array 2312a to the sensing position, and then repeat the aforementioned steps 413~414 to obtain the test spectral image of all pinholes with respect to the point spread function.

If it is confirmed in step 414 that the test spectrum images of all pinholes have been completed, then step 416 is performed to confirm whether the imaging of all depths has been completed. If not, step 417 is performed to change the mirror object to the next depth, and then return to step 413. It should be noted that in step 416, because the test spectrum image corresponding to the new depth is to be obtained, the pinhole of the second spatial filter 231 will be changed to the state shown in FIG. 6B, and then steps 413-415 are performed again to capture images. On the contrary, if it is confirmed in step 416 that all depths have been completed, step 418 is performed to complete all imaging of the point spread function at different depths, and all test spectrum images are stored. It should be noted that the aforementioned embodiment of obtaining the point spread function uses multiple point spread functions obtained from the one-dimensional measurement position on the surface of the object to be measured at the same depth as the point spread functions of different scanning positions. This method can reduce the resources required for calculation and the time required for calculation, thereby speeding up the measurement speed. However, if it is considered that different scanning lines may affect the point spread function and the measurement accuracy due to off-axis, therefore, in another embodiment, between steps 415 and 416, it can further include changing the angle of the galvanometer element 222, adjusting the linear detection light to different scanning positions, and then repeating the aforementioned steps 413-415 to obtain a two-dimensional point spread function, so as to increase the accuracy of measuring the surface morphology.

Returning to FIG. 5A , after obtaining the point spread function image of the pinhole array, step 42 is performed to measure the object to be tested. In an embodiment of step 42, the process is shown in FIG. 7 . In FIG. 7 , step 420 is first performed to adjust the second spatial filter 231 into a slit structure. In this step, because the second spatial filter adopts LCoS, the second spatial filter 231 can be controlled by signal to simulate a slit structure. Afterwards, step 421 is performed to scan the object to be tested and obtain a scanned image of the current scan position. Next, step 422 is performed to obtain a background spectrum image of the current position. In step 422, the calculation processing unit 233 controls the second spatial filter 231 to close the slit, so that the spectrometer senses the background spectrum image at the current position. After that, step 423 is performed to obtain the background-removed spectrum image at the current position. In step 423, the scanned image of step 421 and the background spectrum image of the corresponding scanned position in step 422 are subtracted to obtain the background-removed spectrum image corresponding to the current scanned position.

Then, step 424 is performed to confirm whether the depth scan is completed. If not, step 425 is performed to change to the next scanning position, and then steps 421 to 423 are repeated until all depth scans are completed. It should be noted that in step 425, the scanning position is controlled by changing the angle of the galvanometer element 222 so that the linear dispersion light 92a scans the object to be tested. When step 424 determines that all scans have been completed, step 426 is performed, and the calculation processing unit performs back-convolution calculation on the multiple back-removed spectral images obtained by scanning the surface of the object to be tested in the previous step and the test spectral images corresponding to different pinhole positions obtained in step 41, and then step 427 is performed to determine the surface morphology of the object to be tested based on the response curve of step 40.

In summary, the present invention uses a spatial light modulator, such as liquid crystal on silicon or a combination of filter elements of various sizes, to replace the detection end slit to most directly measure the point spread function of the spectral end that varies with space. By combining the measured spatially varied point spread function with the calculation of the basis deconvolution, the surface image of the object to be tested can be effectively restored to solve the problem of lateral crosstalk.

The above only records the preferred implementation methods or examples of the technical means adopted by the present invention to solve the problem, and is not used to limit the scope of implementation of the present invention. That is, all equivalent changes and modifications that are consistent with the scope of the patent application of the present invention or made according to the scope of the patent of the present invention are covered by the scope of the patent of the present invention.

2: Line scanning color confocal measurement system

20: Light source module

200: Light source

201: Collimator

202: Modulation lens set

202a: Beam expander set

202b: Achromatic lens set

202c, 202d: laminated convex lens

220: Collimator lens set

221:Spectroscopic element

222: Vibration mirror element

223: First Scanning Mirror Set

224:Dispersive objective lens

225: Second scanning lens set

226: Linear polarizer

21: First Space Filter

22:Dispersion module

23: Spectral analysis device

230: Polarization beam splitter element

231: Second Space Filter

232: Spectrometer

233: Operation processing unit

24: Cylindrical concave lens

25: Diffusion film

90:Emits a beam of light

91: Linear detection light

92: Linear filter light

93: Measuring object light

94: Filtering the object measurement light

92a: Linear scattered light

920~922: Linear scattered light

S: Object to be tested

SP: Platform

FP: Front focal plane

Claims (8)

A line scanning color confocal measurement system with global back-convolution surface morphology reconstruction includes: A light source module for generating a linear detection light; A first spatial filter for receiving the linear detection light to form a linear filtered light; A dispersion module for receiving the linear filtered light and projecting the linear filtered light onto an object to be measured to form a measurement light; and A spectrum analysis device for receiving and analyzing the measurement light, the spectrum analysis device further having: A second spatial filter for filtering the measurement light to form a filtered measurement light; A spectrometer for receiving the filtered measurement light to generate corresponding spectral information; and An operation processing unit has a point spread function corresponding to different sensing positions on the second spatial filter at each scanning depth. The operation processing unit receives the spectral information and performs an inverse convolution operation on the point spread function and the spectral information to reconstruct the surface morphology image of the object to be measured. A line scanning color confocal measurement system with global back-convolution surface morphology reconstruction as described in claim 1, wherein the light source module includes: a light source for generating a light beam; a collimator for collimating the light beam to form a collimated light beam; a modulating lens group for modulating the collimated light beam into the linear detection light and modulating the divergence angle of the linear detection light in the linear length direction, and focusing it to the first spatial filter. A line scanning color confocal measurement system with full-domain back-convolution surface morphology reconstruction as described in claim 1, wherein the dispersion module further includes: a galvanometer element, receiving the linear filtered light, the galvanometer element is used to adjust the angle to change the position of the linear filtered light projected onto the object to be measured; a first scanning mirror group, disposed on one side of the galvanometer element, for receiving the linear filtered light; a dispersive objective lens, disposed on one side of the first scanning mirror group, for receiving the linear filtered light, dispersing the linear filtered light to form the dispersed light projected onto the object to be measured and reflected to form the measured object light; and a second scanning mirror group, for receiving the measured object light and guiding the measured object light to the spectral analysis device. A line scanning color confocal measurement system with global backconvolution surface topography reconstruction as described in claim 1, wherein the first spatial filter is a slit structure or a hole structure. A line scanning chromatic confocal measurement system with global backconvolution surface topography reconstruction as described in claim 1, wherein the second spatial filter is a reflective spatial light modulator. A line scanning color confocal measurement system with global backconvolution surface morphology reconstruction as described in claim 5, wherein the second spatial filter allows the object light to pass through without reflecting to the spectrometer to obtain background spectrum information, and the calculation processing unit subtracts the spectrum information from the background spectrum information and performs the backconvolution operation with the point diffusion function. A line scan color confocal measurement system with global backconvolution surface morphology reconstruction as described in claim 1, wherein the second spatial filter further includes: a filter plate having a plurality of filter structures of different sizes or shapes; and a position adjustment element connected to the filter plate, the position adjustment element being used to adjust the position of the filter plate so that one of the filter structures on the filter plate receives the object light. A line scanning color confocal measurement system with global backconvolution surface morphology reconstruction as described in claim 7, wherein the second spatial filter allows the object light to pass through without reflecting to the spectrometer to obtain background spectrum information, and the calculation processing unit subtracts the spectrum information from the background spectrum information and performs the backconvolution operation with the point diffusion function.

Images