High resolution surface measurement method and apparatus based on wide spectrum laser and wavefront coding

Abstract

The application discloses a high-resolution surface measurement method based on wide-spectrum laser and wavefront coding, which modulates the wavefront phase of a wide-spectrum laser light source, focuses the illumination light of different wavelengths on different heights of the normal direction of the surface of a measured object, receives the light radiation in the mirror normal direction of the normal of the surface of the measured object, determines the wavelength with the highest intensity according to the received light radiation, constructs a spectral dark spot distribution, successively illuminates the sample, performs one or more spectral dark spot positioning, determines the accurate central wavelength, and determines the height of the surface of the measured object according to the central wavelength. The light source uses wide-spectrum laser, so that the system light energy utilization is higher and the phase modulation is easier; the phase modulation is used to modulate the illumination light, which is beneficial to reducing the off-axis difference of the system to obtain higher resolution; and the spectral dark spot positioning method is beneficial to obtaining higher imaging positioning accuracy when the imaging signal is weak.

Description

Technical Field

[0001] This invention relates to the field of optical measurement, and more particularly to a high-resolution surface measurement method and apparatus based on broadband laser and wavefront coding. Background Technology

[0002] In the production of precision devices such as optical instruments and semiconductor devices, non-contact three-dimensional surface measurement is important and necessary. Among optical measurement methods, a chromatic aberration measurement scheme can obtain the axial position information of an object's surface without axial scanning, enabling rapid three-dimensional measurement. This scheme is based on the principle that when dispersed light strikes an object's surface, the reflection of different wavelengths corresponds to different axial position information; therefore, the height of the object's surface can be determined by analyzing the detected spectrum.

[0003] Previous research has uncovered a novel method for accurately measuring the height of an object's surface by focusing light of different wavelengths in a modulation manner along the normal direction of the surface being measured. This method effectively avoids errors caused by internal diffuse reflection. However, the accuracy of this method is limited, failing to provide a more reliable and precise solution for measurement tasks requiring high precision. Furthermore, focusing light of different wavelengths onto the normal direction of the measured surface using dispersive components is often difficult and yields unsatisfactory results. Summary of the Invention

[0004] The purpose of this invention is to address the shortcomings of existing technologies by providing a high-resolution non-contact surface measurement method and an apparatus for implementing this method.

[0005] This invention relates to a high-resolution surface measurement method based on broadband laser and wavefront coding. The specific steps of this method are as follows:

[0006] S1. Wavefront phase modulation is performed on the broadband laser source to focus illumination light of different wavelengths at different heights in the normal direction of the surface of the object being measured.

[0007] S2. Receive light radiation from the specular reflection direction of the surface normal of the object being measured, and determine the wavelength λ with the highest intensity based on the received light radiation. c0 ;

[0008] S3, in λ c0 Construct spectral dark spots at k locations nearby, illuminate the sample sequentially, and perform one or more spectral dark spot localization operations to determine the precise center wavelength λ. final ;

[0009] S4. Based on the center wavelength λ final To determine the height of the surface of the object being measured.

[0010] Furthermore, the method for locating spectral dark spots specifically includes:

[0011] The spectral distribution of a dark spot is I(λ-λ c This dark spot is about λ = λ c Symmetric and at λ = λ c The location of this spectral dark spot is defined as λ, where λ is a local minimum. c The width is W;

[0012] During the localization process, at least two points are selected for localization. The selected k spectral dark spot positions are denoted as λ0, λ1, λ2, ..., λk. k-1 The number of signal photons collected at these spectral dark spot locations are n0, n1, n2...n k-1 The total number of photons is N;

[0013] The spectral distribution of the dark spots at these locations is denoted as:

[0014] I i (λ)=I(λ-λ i ), i = 0, 1, 2, ..., k-1;

[0015] The distribution coefficient for each location is defined as follows:

[0016]

[0017] Based on the above information, the probability function (likelihood function) of the location distribution can be calculated:

[0018]

[0019] The spectral position λ with the highest probability distribution is obtained by the maximum likelihood estimation method. MLE :

[0020]

[0021] This expression has two solutions when k=2 takes two points, and only has a unique solution when k=3.

[0022] The corresponding positioning accuracy σ CRB for:

[0023]

[0024] Furthermore, the specific operation method for single-spectral dark spot localization is as follows:

[0025] For the determined center wavelength λ c0 Around this wavelength, k spectral dark spot positions are selected symmetrically with a maximum width of L1, denoted as λ0, λ1, λ2, ..., λk. k-1The structural centers are located at λ0, λ1, λ2……λ k-1 The spectral dark spots were used to illuminate the samples in turn, and the collected signal intensities were denoted as F0, F1, F2...F k-1 The signal intensity F is proportional to the number of photons n; the center wavelength λ is calculated based on the signal intensity. c1 and positioning accuracy σ CRB1 For a single positioning λ c1 That is, the final λ final .

[0026] Furthermore, the specific operational method for locating dark spots in multiple spectra is as follows:

[0027] For the determined center wavelength λ c0 Around this wavelength, k spectral dark spot positions are selected symmetrically with a maximum width of L1, denoted as λ0, λ1, λ2, ..., λk. k-1 The structural centers are located at λ0, λ1, λ2……λ k-1 The dark spots were used to illuminate the samples in turn, and the collected signal intensities were recorded as F0, F1, F2...F k-1 The signal intensity F is proportional to the number of photons n; the center wavelength λ is calculated based on the signal intensity. c1 and positioning accuracy σ CRB1 ;

[0028] With λ c1 The center wavelength for the next dark spot localization, based on the localization accuracy σ CRB1 Set the maximum width for the next scan, for example, with L2 = 3σ. CRB1 To determine the maximum width for the next scan, repeat the above operation iteratively for dark spot localization, keeping the number of selected spectral dark spot locations k constant; the iteration terminates when a specified number of iterations or the σ obtained in the s-th iteration is reached. CRBs Stop when the value is less than the accuracy requirement; the iteration terminates when λ... cs That is, the final λ final .

[0029] Furthermore, the I(λ) to be substituted into the calculation is the dark spot spectral distribution, which needs to be obtained through pre-calibration. The dark spot spectral distribution is a spectral distribution curve that has a minimum value at the center, is symmetrical about the center, and has monotonic spectral curves on both sides of the center; the larger the ratio of the minimum value at the center of the spectral distribution to the maximum value of the entire distribution, the worse the positioning accuracy under the same signal strength; a spectral distribution curve with poor symmetry about the center can still be used for positioning, but it will affect the positioning accuracy.

[0030] Furthermore, the selected spectral dark spot location is within (λ). c0 -L1 / 2,λ c0Within the range of +L1 / 2], k points are selected at equal intervals, λ0=λ c0 -L1 / 2,λ k-1 =λ c0 +L1 / 2;

[0031] The maximum width L1 selected initially is chosen based on requirements. For a single spectral dark spot localization operation, the larger L1 is, the higher the localization accuracy σ. CRB1 The larger the value, the larger L1 is, and the higher the positioning accuracy σ. CRB1 The smaller the value, the more iterations are required to achieve the same positioning accuracy for multiple spectral dark spot localization operations.

[0032] Furthermore, the k constructed spectral dark spots sequentially illuminate the sample in turn, and the exposure method is to adopt a fixed detection exposure time or to adopt an extremely short exposure time and expose multiple times;

[0033] The fixed detection exposure time is used to detect all k positions once with a set long fixed exposure time. The sum of the light intensity of the k detection results is used as the basis for calculating the total photon number N, or the sum of the photon number of the k detection results is used as the total photon number N.

[0034] The method of using extremely short exposure times and multiple exposures involves repeatedly scanning the k points in turn with a set extremely short fixed exposure time until the total number of accumulated photons reaches a fixed value N, which is used to achieve higher positioning accuracy faster when the signal is weaker.

[0035] Furthermore, S4 specifically involves obtaining the mapping relationship between light wavelength and height through calibration or calculation, thereby determining the relationship based on the measured λ. final Calculate the height of the sample to be tested.

[0036] According to a second aspect of the present invention, a high-resolution surface measurement device based on broadband laser and wavefront coding is provided, the device comprising: a broadband laser source, a spectral dark spot modulation unit, a phase modulation unit, a phase recovery unit, a detector, and a signal processing unit;

[0037] The spectral dark spot modulation unit is a device for constructing spectral dark spots on the spectral intensity distribution; the unit modulates the spectral distribution of the incident illumination source to construct spectral dark spots on the spectral intensity distribution at different locations;

[0038] The phase modulation unit is a device for performing wavefront phase encoding on the light source; by performing wavefront phase encoding on the light source, illumination light of different wavelengths is focused at different heights in the normal direction of the surface of the object being measured.

[0039] The phase recovery unit receives light radiation from the object being measured and recovers the modulated phase into parallel light that enters the detector.

[0040] The detector is a detection element or a detection module composed of multiple elements for detecting the spectral distribution of the detection signal;

[0041] The signal processing unit analyzes the intensity of the detected probe light at different wavelengths and controls the spectral dark spot modulation unit based on the detection results.

[0042] Furthermore, when the core detection element in the detector uses a single-point detection element that only provides the number of detected photons without detecting the spectral signal, a detection module composed of multiple elements is used to realize the function of the detector. In the detection module, a 50% beam splitter is used to split the signal after the phase recovery unit into two beams of equal intensity. A filter with a monotonic transmission spectral function is added to one of the beams. The two beams are respectively incident on two single-point detection elements with identical parameters: a first detection element and a second detection element. The signals collected by the first and second detection elements are compared; the ratio signal represents the transmittance of the wavelength with the highest intensity, thus obtaining the wavelength λ with the highest spectral intensity. c0 .

[0043] The beneficial effects of this invention are:

[0044] The light source of this invention uses broadband laser, which is a coherent light source with continuous wavelength, making the system's light energy utilization rate higher and easier to phase modulate;

[0045] This invention uses phase modulation to modulate the illumination light, focusing illumination light of different wavelengths at different heights along the normal direction of the surface of the object being measured. Compared with existing dispersive elements, this operation is easier to implement, requires no customization, has lower cost, is easier to adjust the system, and helps to reduce the off-axis phase difference of the system to obtain higher resolution and signal-to-noise ratio.

[0046] This invention introduces a method for spectral dark spot localization, which is beneficial for obtaining higher imaging localization accuracy when the imaging signal is weak, and can reduce the impact on localization accuracy caused by the inconsistency of reflection or transmission efficiency of light of different wavelengths when passing through the device. Attached Figure Description

[0047] Figure 1 This is a schematic diagram of the spectral distribution of the dark spot.

[0048] Figure 2 This is a schematic diagram of the overall device.

[0049] Figure 3 This describes the device architecture that can be used when the detector is a single-point detector;

[0050] Figure 4 A phase modulation unit that uses a reflective phase modulation device;

[0051] Figure 5 A phase modulation unit that uses a transmissive phase modulation device;

[0052] Figure 6 A phase recovery unit using a reflective phase modulation device;

[0053] Figure 7 This is a phase recovery unit that uses a transmission-type phase modulation device. Detailed Implementation

[0054] The specific embodiments of the present invention will be further described in detail below with reference to the accompanying drawings.

[0055] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0056] The specific steps of the method of the present invention are as follows:

[0057] S1. Wavefront phase modulation is performed on the broadband laser source to focus illumination light of different wavelengths at different heights in the normal direction of the surface of the object being measured.

[0058] S2. Receive light radiation from the specular reflection direction of the surface normal of the object being measured, and determine the wavelength λ with the highest intensity based on the received light radiation. c0 ;

[0059] S3, in λ c0 Construct spectral dark spots at k locations nearby, illuminate the sample sequentially, and perform one or more spectral dark spot localization operations to determine the precise center wavelength λ. final ;

[0060] S4. Based on the center wavelength λ final To determine the height of the surface of the object being measured.

[0061] The principle of the spectral dark spot localization method is as follows:

[0062] Consider a dark spot with a spectral distribution of I(λ-λ). c This dark spot is about λ = λ c Symmetric and at λ = λ c There is a local minimum at λ. We define the location of this dark spot as λ. c The width is W. For example... Figure 1 The diagram shows two possible spectral distributions of a spectral dark spot, illustrating the distribution of light intensity by wavelength within the spectral dark spot. Figure 1 The left figure shows the spectral distribution of a bimodal curve with a minimum value at the center. Figure 1The right figure shows the spectral distribution of the curve that is concave only in the center.

[0063] During the localization process, multiple points (at least two) are selected for localization. The selected k dark spot locations are denoted as λ0, λ1, λ2…λ k-1 Let the number of signal photons collected at these locations be defined as n0, n1, n2...n k-1 The total number of photons is N.

[0064] The spectral distribution of the dark spots at these locations is denoted as:

[0065] I i (λ)=I(λ-λ i ), i = 0, 1, 2, ..., k-1. (1) The distribution coefficient for each location is defined as follows:

[0066]

[0067] Based on the above information, the probability function (likelihood function) of the location distribution can be calculated:

[0068]

[0069] The spectral position λ with the highest probability distribution is obtained by the maximum likelihood estimation method (MLE). MLE :

[0070]

[0071] Note that this formula has two solutions when k=2 takes two points, and only has a unique solution when k=3.

[0072] The corresponding positioning accuracy σ CRB for:

[0073]

[0074] The specific operation method for single-spectral dark spot localization is as follows:

[0075] For the determined center wavelength λ c0 Around this wavelength, k spectral dark spot positions are selected symmetrically with a maximum width of L1, denoted as λ0, λ1, λ2...λ k-1 The structural centers are located at λ0, λ1, λ2……λ k-1 The spectral dark spots were used to illuminate the samples in turn, and the collected signal intensities were denoted as F0, F1, F2...F k-1 The signal intensity F is proportional to the number of photons n. The center wavelength λ is calculated based on this set of signal intensities. c1 and positioning accuracy σ CRB1 For a single positioning λc1 That is, the final λ final .

[0076] The specific operation method for locating dark spots in multiple spectra is as follows:

[0077] For the determined center wavelength λ c0 Around this wavelength, k dark spot positions are selected symmetrically with a maximum width of L1, denoted as λ0, λ1, λ2...λ k-1 The spectral dark spot modulation unit is constructed with the center located at λ0, λ1, λ2……λ k-1 The dark spots were used to illuminate the samples in turn, and the collected signal intensities were recorded as F0, F1, F2...F k-1 The signal intensity F is proportional to the number of photons n. The center wavelength λ is calculated based on this set of signal intensities. c1 and positioning accuracy σ CRB1 With λ c1 The center wavelength for the next dark spot localization, based on the localization accuracy σ CRB1 Set the maximum width for the next scan, for example, with L2 = 3σ. CCB1 To determine the maximum width for the next scan, repeat the above steps to iterate the dark spot localization, keeping the number of selected dark spot locations, k, constant. The iteration terminates when a specified number of iterations or the σ obtained on the s-th iteration is reached. CRBs Stop when the value is less than the accuracy requirement; the iteration terminates when λ... cs That is, the final λ final .

[0078] In this calculation, I(λ) is the spectral distribution of the dark spot, which needs to be obtained through pre-calibration. This spectral distribution needs to be a spectral distribution curve with a minimum value at the center and should be as symmetrical about the center as possible. The spectral distribution can be a bimodal curve with a minimum value at the center or a curve that is concave only at the center (e.g., ...). Figure 1 As long as the spectral curves on both sides of the center are monotonic, the positioning effect of this method can be achieved. A minimum value of 0 at the center is optimal; however, it can still be used for positioning even if it's not zero. The larger the ratio of the minimum value at the center of the spectral distribution to the maximum value of the entire distribution (defined as dark spot contrast), the worse the positioning accuracy will be for the same signal strength. Spectral distribution curves with poor central symmetry can still be used for positioning, but this will affect the positioning accuracy.

[0079] The selected dark spot location is in [λ]. c0 -L1 / 2,λ c0 Within the range of +L1 / 2], k points are selected at equal intervals, λ0=λ c0 -L1 / 2,λ k-1 =λ c0+L1 / 2. For one-dimensional spectra, k = 2 or 3 is generally chosen. When k = 2, the scanning speed is faster, but a non-unique solution may occur, in which case λ0 = λ c0 -L1 / 2,λ1=λ c0 +L1 / 2. k=3 guarantees the uniqueness of the location, at which point λ0=λ c0 -L1 / 2,λ1=λ c0 ,λ2=λ c0 +L1 / 2.

[0080] The maximum width L1 selected initially can be chosen according to requirements. For a single spectral dark spot localization operation, the larger L1 is, the higher the localization accuracy σ. CRB1 The larger the value, the larger L1 is, and the higher the positioning accuracy σ. cRB1 A smaller L1 value results in a higher accuracy, but an excessively small L1 value may cause the target to be lost during dark spot localization. For multiple spectral dark spot localization operations, a larger L1 value means more iterations are required to achieve the same localization accuracy, but an excessively small L1 value may also cause the target to be lost during the first dark spot localization.

[0081] The structuring centers are located at λ0, λ1, λ2……λ k-1 The process involves sequentially illuminating the sample with spectral dark spots. One approach is to use a fixed exposure time, probing each of the k locations once, and using the sum of the light intensity from the k probes as the basis for calculating the total photon count N. Alternatively, an extremely short exposure time can be chosen, sequentially scanning the k points until the cumulative light intensity reaches a fixed value N. The latter method can achieve higher positioning accuracy more quickly even with weaker signals.

[0082] To implement the method described in this invention, the imaging device provided by this invention is as follows: Figure 2 As shown, it includes an illumination modulation module 100 and a detection modulation module 120 that are separate from each other. The illumination modulation module includes a broadband laser source 101, a spectral dark spot modulation unit 102, and a phase modulation unit 103; the detection modulation module includes a phase recovery unit 121 and a detector 122; the signal detection unit 130 is connected to the broadband laser source 101, the spectral dark spot modulation unit 102, and the detector 122.

[0083] During the detection process, the laser emitted from the broadband laser source 101 passes through the spectral dark spot modulation unit 102 (which initially performs no modulation) and enters the phase modulation unit 103. The phase modulation unit 103 achieves phase modulation by editing the wavefront of the illumination light, focusing light of different wavelengths from the broadband laser source 101 at different heights along the normal direction of the surface of the sample 110 under test. The illumination modulation module 100 focuses the illumination light at different heights along the normal direction of the surface of the sample 110 under test. Light with wavelengths at the same height as the sample 110 under test is reflected by the surface and enters the detection modulation module 120. The phase recovery unit 121 is symmetrical to the phase modulation unit 103 about the normal of the surface under test and uses a wavefront editing device symmetrical to the phase modulation unit 103. After entering the phase recovery unit 121, the detection signal is edited into parallel light and enters the detector 122. The detector 122 converts the collected spectral signal into an electrical signal with a spectral distribution and transmits it to the signal processing unit 130. The signal processing unit 130 analyzes the collected detection signal spectrum to obtain the wavelength λ with the highest spectral intensity. c0 .

[0084] For the determined center wavelength λ c0 Three dark spot locations are selected symmetrically around this wavelength position, with a maximum width of L1, denoted as λ0 = λ c0 -L1 / 2,λ1=λ c0 ,λ2=λ c0 +L1 / 2. The signal processing unit 130 controls the spectral dark spot modulation unit 102 to construct dark spots centered at λ0, λ1, and λ2. The sample is illuminated with the same exposure time t, and three spectral distribution signals are collected by the detector 122 respectively. The integral of the spectral intensity distribution collected for each acquisition (i.e., the sum of the intensities at different wavelengths) is denoted as F0, F1, and F2. The signal intensity F is proportional to the number of photons n, and the proportionality coefficient can be obtained by pre-calibration. Substituting the photon numbers n0, n1, and n2 converted from these signal intensities into equation (3), the center wavelength λ is calculated. c1 and positioning accuracy σ CRB1 .

[0085] For a single positioning λ c1 That is, the final λ final If iterative improvements in positioning accuracy are desired, then λ is used. c1 The center wavelength for the next dark spot localization is L2 = 3σ. CRB1 To determine the maximum width for the next scan, the above operation is repeated at three dark spot locations to iterate the dark spot localization. The iteration can terminate after a specified number of iterations (s), or at the σ value obtained on the s-th iteration. CRBs Stop when the value is less than the accuracy requirement; the iteration terminates when λ... cs That is, the final λfinal .

[0086] The phase modulation unit 103 focuses light of different wavelengths onto a specific height in the normal direction of the surface of the sample 110. We can obtain the mapping relationship between light wavelength and height through calibration or calculation, and then, based on the measured λ... final The height h of the sample to be tested is calculated.

[0087] Among them, the spectral dark spot modulation unit 102 is an acousto-optic tunable filter (AOTF), a tunable filter, or other device that can realize the construction of spectral dark spots on the spectral intensity distribution.

[0088] Among them, the phase editing elements in the phase modulation unit 103 and the phase recovery unit 121 are spatial light modulators (SLMs), digital micromirror arrays (DMDs), reflective or transmissive Fresnel masks or lenses, etc., which can realize wavefront phase encoding of light sources.

[0089] The detector can be a detector with spectral resolution, such as a color CCD or a spectrometer.

[0090] Figure 3 The system described above presents a system design that can implement this method when the core detection element in the detector uses a single-point detection element, such as a PMT or APD, which only provides the number of photons detected and does not detect the spectral signal. In this case, we use a 50% beam splitter 201 to split the signal after the phase recovery unit 121 into two beams of equal intensity. A filter 202 with a monotonic transmission spectral function is added to one of the beams. The two beams are respectively incident on two single-point detection elements, the first detection element 203 and the second detection element 204, which have identical parameters. Since the transmission spectral function of the filter 202 is monotonic, that is, the transmittance has a one-to-one correspondence with the wavelength, we compare the signals acquired by the first detection element 203 and the second detection element 204. This ratio signal is the transmittance of the wavelength with the highest intensity. Therefore, we can obtain the wavelength λ with the highest spectral intensity. c0 .

[0091] For the determined center wavelength λ c0 Three dark spot locations are selected symmetrically around this wavelength position, with a maximum width of L1, denoted as λ0 = λ c0 -L1 / 2,λ1=λ c0 ,λ2=λ c0+L1 / 2. The signal processing unit 130 controls the spectral dark spot modulation unit 102 to construct dark spots centered at λ0, λ1, and λ2. The sample is illuminated with the same exposure time t, and three photon count signals, denoted as n0, n1, and n2, are collected by the second detection element 204. Substituting these photon counts n0, n1, and n2 into equation (3), the center wavelength λ is calculated. c1 and positioning accuracy σ CRB1 For a single positioning λ c1 That is, the final λ final If iterative improvements are made to increase positioning accuracy, the λ value at the end of the iteration... cs That is, the final λ final .

[0092] If better imaging results are desired when the imaging signal is weak, then for a given center wavelength λ c0 Three dark spot locations are selected symmetrically around this wavelength position, with a maximum width of L1, denoted as λ0 = λ c0 -L1 / 2,λ1=λ c0 ,λ2=λ c0 +L1 / 2. The signal processing unit 130 controls the spectral dark spot modulation unit 102 to construct dark spots centered at λ0, λ1, and λ2. The three dark spot positions are exposed alternately with a short exposure time t0. At the same time, the second detection element 204 collects three photon count signals and accumulates them, denoted as n0, n1, and n2. The exposure is carried out alternately until the sum of the accumulated photon counts n0, n1, and n2 reaches the set fixed total number of photons N. Substituting this set of photon counts n0, n1, and n2 into equation (3), the center wavelength λ is calculated. c1 and positioning accuracy σ CRB1 For a single positioning λ c1 That is, the final λ final If iterative improvements are made to increase positioning accuracy, the λ value at the end of the iteration... cs That is, the final λ final The extremely short exposure time is an order of magnitude shorter than the exposure time for fixed detection.

[0093] Figure 4This is a phase modulation unit using a first reflective phase modulation device 302. The first reflective phase modulation device 302 can be a reflective element such as an SLM or DMD that can achieve wavefront phase encoding. After passing through the spectral dark spot modulation unit 102, the broadband laser source 101 needs to be incident on the phase modulation unit 103 as parallel light. In the phase modulation unit 103, the incident light hits the reflective phase modulation device 302 via the first reflecting mirror 301, and the modulated light is incident on the first objective lens 304 via the second reflecting mirror 303, and is focused by the first objective lens 304 at different heights in the normal direction of the surface of the sample 110 under test. In the phase modulation unit 103, the function of the first reflecting mirror 301 and the second reflecting mirror 303 is to allow the incident light to enter the first reflective phase modulation device 302 at a small angle as much as possible.

[0094] Figure 5 This is a phase modulation unit that uses a transmissive phase modulation device. The transmissive phase modulation device can be a transmissive element such as a Fresnel zone plate that can achieve wavefront phase encoding. After passing through the spectral dark spot modulation unit 102, the broadband laser source 101 needs to be incident as parallel light on the phase modulation unit 103, hitting the second reflective phase modulation device 401. The modulated light is then incident on the second objective lens 402 and focused by the second objective lens 402 at different heights in the normal direction of the surface of the sample 110 under test.

[0095] Figure 6 Phase recovery unit 121 uses a third reflective phase modulator 503. The third reflective phase modulator 503 can be a reflective element such as an SLM or DMD that can achieve wavefront phase encoding. Illumination light reflected from the surface under test enters the phase recovery unit 121, is collected by the third objective lens 501, and then strikes the third reflective phase modulator 503 via the third mirror 502. After modulation, it is incident on the converging lens 505 via the fourth mirror 504 and finally focused onto the detector 122. In the phase modulation unit 121, the function of the third mirror 502 and the fourth mirror 504 is to ensure that the incident light enters the third reflective phase modulator 503 at a small angle as much as possible. The third reflective phase modulator 503 is configured to be symmetrical with the phase modulator in the phase modulation unit 103, regardless of whether the phase modulator is reflective or transmissive.

[0096] Figure 7Phase recovery unit 121 uses a transmissive phase modulator. The transmissive phase modulator can be a transmissive element such as a Fresnel zone plate that can achieve wavefront phase encoding. Illumination light reflected from the surface under test enters phase recovery unit 121, is collected by fourth objective lens 601, and then strikes fourth reflective phase modulator 602. After modulation, it is incident on converging lens 603 and finally focused onto detector 122. Fourth reflective phase modulator 602 is configured to be symmetrical with the phase modulator in phase modulation unit 103, regardless of whether the phase modulator is reflective or transmissive.

[0097] The above diagram is for illustrative purposes only. The lenses, mirrors, and polarizers required for beam expansion, deflection, and polarization modulation in the actual system are not shown. They need to be adjusted according to the actual system requirements and application scenarios.

[0098] The above embodiments are used to explain and illustrate the present invention, but not to limit the present invention. Any modifications and changes made to the present invention within the spirit and scope of the claims shall fall within the protection scope of the present invention.

Claims

1. A high-resolution surface measurement method based on wide-spectrum laser and wavefront coding, characterized by, The method comprises the following specific steps: S1, wavefront phase modulation is performed on the wide spectrum laser light source, and illumination light of different wavelengths is focused on different heights in the normal direction of the surface of the measured object; S2, receiving light radiation from the direction of specular reflection of the surface normal of the object being measured, determining the wavelength with the highest intensity from the received light radiation ​ S3、According to the wavelength with the highest intensity Selecting k positions to construct a spectral dark spot, sequentially illuminating the sample, and performing one or more spectral dark spot positioning to determine the accurate center wavelength The spectral dark spot positioning method is specifically as follows: The spectral distribution of a spectral dark spot is This dark spot is symmetric about and has a minimum at , i.e. the position of this spectral dark spot is defined by and has a width of ; In the positioning process, at least 2 points are selected for positioning, and for the selected k spectral dark spot positions, denoted as The number of signal photons collected at these spectral dark spot positions is respectively The total number of photons is N; The spectral distribution of the dark spots at these positions is denoted as: , i = 0,1,2……k-1; The distribution coefficient of each position is defined as: , i = 0,1,2……k-1; Then, the probability function of the positioning distribution is calculated according to the distribution coefficient of each position: Obtaining the spectral position with the highest distribution probability by maximum likelihood estimation : This formula has two solutions when There are two solutions when two points are taken, There is only one solution when Corresponding positioning accuracy is: S4. determining the height of the surface of the object being measured based on the center wavelength S4. determining the height of the surface of the object being measured based on the center wavelength 2. The high-resolution surface measurement method based on wide-spectrum laser and wavefront coding according to claim 1, characterized in that, The specific operation method of single spectral dark spot positioning is as follows: For the determined center wavelength , the maximum width of the spectrum dark spot symmetrical around the wavelength position is Select k spectrum dark spot positions, denoted as , construct a spectrum dark spot with the center at , illuminate the sample in turn, and collect the collected signal intensity, denoted as The signal intensity F is proportional to the number of photons n; according to the signal intensity, the center wavelength and the positioning accuracy are calculated; for single positioning , the final .

3. The high-resolution surface measurement method based on wide-spectrum laser and wavefront coding according to claim 1, characterized in that, The specific operation method of multiple spectral dark spot positioning is as follows: For the determined center wavelength , the maximum width of the symmetric wavelength position is Select k spectral dark spot positions, denoted as , construct a dark spot with the center at , illuminate the sample in turn, and collect the collected signal intensity, denoted as The signal intensity F is proportional to the number of photons n; according to the signal intensity, the center wavelength and the positioning accuracy are calculated. by The center wavelength for the next dark spot localization, based on localization accuracy. Set the maximum width for the next scan, repeat the above operation, and iterate the dark spot localization. The number of selected spectral dark spot positions, k, remains unchanged. The iteration terminates when a specified number of iterations or the result obtained in the s-th iteration is reached. Stop when the value is less than the precision requirement, at the time of iteration termination. That is, the final .

4. The high-resolution surface measurement method based on wide-spectrum laser and wavefront coding according to claim 1, characterized in that, Dark spot spectral distribution Needs to be obtained by pre-calibration; The spectral distribution of the dark spots needs to have a spectral distribution curve with a minimum value at the center, which is symmetric about the center, and the spectral curve on both sides of the center is monotonic; the greater the ratio of the minimum value at the center to the maximum value of the entire distribution, the worse the positioning accuracy under the same signal intensity; spectral distribution curves that are not well symmetric about the center can still be used for positioning, but will affect the positioning accuracy.

5. The high-resolution surface measurement method based on wide-spectrum laser and wavefront coding according to claim 1, characterized in that, The selected spectral dark spot positions are k points selected at equal intervals in the range of 400 to 700 nm, , , ; wherein the maximum width selected for the first time According to the requirement, for single spectrum dark spot positioning operation, The larger the value is, the higher the positioning accuracy The larger the value is, the higher the positioning accuracy The smaller the value is, the higher the positioning accuracy The smaller the value is, the higher the positioning accuracy; for multiple spectrum dark spot positioning operation, The larger the value is, the more the iteration times required to reach the same positioning accuracy.

6. The high-resolution surface measurement method based on wide-spectrum laser and wavefront coding according to claim 1, characterized in that, k spectral dark spots are constructed to illuminate the sample in turn, and the exposure mode is to adopt a fixed detection exposure time or to adopt an extremely short exposure time and multiple exposures; The fixed detection exposure time is that all k positions are detected once with a set fixed exposure time, and the total light intensity of the k detection results is used as the basis for calculating the total photon number N, or the total photon number of the k detection results is used as the total photon number N; The extremely short exposure time and multiple exposures are to repeatedly scan the k points in turn with a set extremely short fixed exposure time until the total accumulated photon number reaches a fixed value N, which is used to achieve higher positioning accuracy faster in the case of weaker signals, and the extremely short exposure time is one order of magnitude smaller than the fixed detection exposure time.

7. The high-resolution surface measurement method based on wide-spectrum laser and wavefront coding according to claim 1, characterized in that, S4 is specifically obtaining the mapping relationship between the light wavelength and the height by calibration or calculation, so as to obtain the height of the sample according to the measured The height of the sample to be measured is calculated.

8. A high resolution surface measurement device based on a broad spectrum laser and wavefront coding for implementing the method of any of claims 1 to 7, characterized in that, The device comprises a wide spectrum laser light source, a spectral dark spot modulation unit, a phase modulation unit, a phase recovery unit, a detector, and a signal processing unit; The spectral dark spot modulation unit is a device for constructing spectral dark spots on the spectral intensity distribution; this unit modulates the spectral distribution of the incident illumination light source to construct spectral dark spots on the spectral intensity distribution at different positions; The phase modulation unit is a device for performing wavefront phase coding on the light source; the wavefront phase coding is performed on the light source so that illumination light of different wavelengths is focused on different heights in the normal direction of the surface of the measured object; The phase recovery unit receives optical radiation from the measured object and recovers the modulated phase into parallel light entering the detector; The detector is a detection element or a detection module composed of multiple elements for detecting the detected signal spectral distribution; The signal processing unit analyzes the intensity of the detected light of different wavelengths and controls the spectral dark spot modulation unit according to the detection results.

9. The high resolution surface measurement device based on a broad spectrum laser and wavefront coding of claim 8, wherein, When the core detection element in the detector uses a single-point detection element that does not detect a spectrum signal but only gives the number of detected photons, a detection module composed of multiple elements is used to realize the function of the detector; the signal after the phase recovery unit is divided into two beams with equal light intensity using a 50% beam splitter, a filter piece with a monotonous transmission spectrum function is added in one of the two beams, and the two beams of light are respectively incident on two single-point detection elements, a first detection element and a second detection element, which are completely the same in parameters; the ratio of the signals collected by the first detection element and the second detection element is calculated, and the ratio signal is the transmittance of the wavelength with the highest intensity, so that the wavelength with the highest spectral distribution intensity is obtained .

Images