Optical interference phase extraction method based on principal component analysis and vector inverse orthogonalization

Through principal component analysis and vector inverse orthogonalization optical interference phase extraction method, the problem of low accuracy and time-consuming in optical phase shift interference measurement is solved, and the robustness of background fluctuations and modulation fluctuations in interferometric images is achieved. , significantly improves calculation accuracy and speed, suitable for precision measurement.

CN118980435AActive Publication Date: 2024-11-19XI AN JIAOTONG UNIV
View PDF 5 Cites 0 Cited by

Patent Information

Application Number
CN202411205636.X
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-08-30
Publication Date
2024-11-19
Estimated Expiration
2044-08-30

AI Technical Summary

Technical Problem

There are problems in the existing optical phase shift interference measurement technology that are not high, easily error-producing, and time-consuming. Especially when background fluctuations and modulation nonlinear fluctuations in the interference image, it is difficult to accurately extract the phase of the object to be measured. information.

Method used

By using the optical interference phase extraction method based on principal component analysis and vector inverse orthogonalization, the initial basis vector is solved by treating the light intensity value of the interference image as a multi-dimensional light intensity vector, and the principal component analysis method and vector inverse orthogonalization are used to solve the initial basis vector and calculate the The phase shift between the phases of the interfering map and the wrapping phase of the object surface height, and the effect of background light intensity fluctuations is eliminated through iterative correction.

Benefits of technology

It realizes robustness of background light intensity fluctuations and modulation fluctuations in interference images, significantly improving calculation accuracy and speed. It is especially suitable for precise measurement of samples to be measured with high surface machining accuracy, with less calculation error than The AIA algorithm takes only 20% of the AIA algorithm.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118980435A_ABST
    Figure CN118980435A_ABST
Patent Text Reader

Abstract

The invention discloses an optical interference phase extraction method based on principal component analysis and vector inverse orthogonalization, which comprises the following steps of: regarding a light intensity value of the same pixel point in an interference image set as a multi-dimensional light intensity vector, and regarding light intensity vectors of different pixel points as a result which is linearly expressed by a group of unknown initial basis vectors according to a double-beam interference light intensity formula. The real phase shift between the interference images is obtained by solving the unknown initial basis vector, and finally the wrapped phase of the object surface height is solved; during calculation, a group of standard orthogonal basis of an interference light intensity vector is solved through a principal component analysis method, and the first two principal components of the standard orthogonal basis are regarded as the orthogonalization result of an unknown initial basis vector; then setting constraint conditions to establish an equation set, obtaining an initial basis vector through vector inverse orthogonal calculation, calculating a real phase shift amount between interference images, substituting the real phase shift amount into a linear overdetermined equation set constructed according to a double-beam interference light intensity theory, and solving a least square solution of a highly wrapped phase of an object surface;
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention belongs to the technical field of optical precision measurement, relates to optical phase shift interference measurement, and particularly to an optical interference phase extraction method based on principal component analysis and vector inverse orthogonalization. Background Art

[0002] Optical phase-shift interferometry is a precision measurement method based on the principle of light wave interference, which is widely used in industrial manufacturing and medical diagnosis. The basic principle is that when two coherent light beams meet at a certain point in space, the phase relationship of the coherent light waves will cause the light intensity to change, thus forming an interference pattern. The core of optical phase-shift interferometry is to use optical phase extraction technology to calculate the height phase of the surface of the object being measured.

[0003] The traditional phase-shift interferometric phase extraction technology requires a fixed step size for a single phase shift. However, the current interferometric microscopy measurement system generally uses a micro-piezoelectric displacer installed on the interferometric objective lens to perform interferometric phase shifting operations. This micro-piezoelectric displacer generally has motion errors of several to dozens of nanometers during movement, which leads to phase extraction calculation errors. In recent years, there are two major development trends in the phase extraction technology of optical phase-shift interferometry. One is to improve the system measurement efficiency by reducing the number of interference images involved in the calculation. Such algorithms include two-step phase shift algorithms such as the Glan-Schmidt orthogonalization method. This type of method has a fast measurement speed, but is sensitive to noise and has lower calculation accuracy than the multi-step phase shift method. Therefore, it is only suitable for fast online measurement of samples. Second, some multi-step blind phase shift technologies that do not require phase shifts have emerged. The most representative of these algorithms are the principal component analysis algorithm and the advanced iterative algorithm that do not require iteration. Although the non-iterative algorithms represented by the principal component analysis phase extraction algorithm have a fast calculation speed, the calculation accuracy is often not as good as the iterative method and it is difficult to determine the global phase sign of the object being measured. Although iterative algorithms can often obtain very accurate solutions, they require multiple iterations and take a long time to solve. When there are large background fluctuations and nonlinear fluctuations in the modulation index in the interference image, non-negligible calculation errors will occur. In addition, some algorithms have requirements on the number of fringes in the interference pattern. When the height phase distribution of the object surface is less than one period, large calculation errors will occur, which is not conducive to the measurement of samples with flat surfaces. Therefore, there is an urgent need for a multi-step blind phase-shifting interferometry phase extraction technology that has extremely high robustness, no requirements on the number of fringes in the interference image, and both calculation speed and calculation accuracy.

[0004] The above technical problems are the main difficulties faced in current optical interference phase extraction technology. Summary of the invention

[0005] In order to overcome the shortcomings of the above-mentioned prior art, the purpose of the present invention is to provide an optical interferometry phase extraction method (PCAVR) based on principal component analysis and vector inverse orthogonalization, so as to solve the problems of low accuracy, easy to produce errors, and long time consumption in the current optical interferometry phase extraction technology.

[0006] In order to achieve the above object, the present invention regards the light intensity value of the same pixel point in the interference image set as a multidimensional light intensity vector linearly represented by a set of unknown initial basis vectors, and the real phase shift between the unknown initial basis vectors and the interference image is a trigonometric function relationship. The unknown initial basis vectors are solved by principal component analysis and vector inverse orthogonalization, and the phase shift between the interference patterns and the wrapped phase of the object surface height are calculated.

[0007] The specific technical solution of the present invention can be described as follows:

[0008] An optical interference phase extraction method based on principal component analysis and vector inverse orthogonalization comprises the following steps:

[0009] Step 1, reconstructing the interference image data according to the requirements of the principal component analysis method to obtain a light intensity matrix I, and the number of interference images is at least 4;

[0010] Step 2: Calculate the covariance matrix C of the light intensity matrix I, and perform principal component analysis to obtain the two principal element vectors p1 and p2 corresponding to the two largest eigenvalues ​​λ1 and λ2;

[0011] Step 3: According to the constraint conditions of the basis vectors, the orthogonal principal element vectors p1 and p2 are vector-inverse orthogonalized to solve the initial basis vectors V1 and V2;

[0012] Step 4: Perform an inverse tangent operation on the initial basis vectors V1 and V2 to obtain the wrapped phase shift between the interference images; perform one-dimensional phase unwrapping and phase flip alignment on the wrapped phase shift between the interference images to obtain the true phase shift δ between the interference images n ;

[0013] Step 5, constructing an overdetermined linear equation system containing the height phase information of the object to be measured, and finding its least square solution to obtain the surface height phase information of the object to be measured;

[0014] Step 6, analyzing the eigenvalues ​​obtained by principal component analysis to determine whether the phase calculation result is reliable. If the calculation result is reliable, the surface height phase information of the measured object is output and the calculation is terminated;

[0015] Step 7: When the calculation result is judged to be unreliable, the light intensity matrix I is corrected using the solution in step 5, and steps 2 to 6 are repeated for iterative calculation until the algorithm converges or reaches the set maximum number of iterations.

[0016] Compared with the prior art, the present invention has the following beneficial effects:

[0017] The optical interference phase extraction method based on principal component analysis and vector inverse orthogonalization of the present invention is not only completely insensitive to modulation fluctuations in the interference image, but also can judge the quality of the interference image through multiple thresholds. Even if the interference image used for phase extraction has background light intensity fluctuations, the original interference image can be continuously corrected by iteration to eliminate the influence of the background light intensity fluctuations between pixels in the interference image on the phase extraction, thereby accurately calculating the height phase information of the measured object. When the interference image quality is high and there is no obvious background light intensity fluctuation, the method of the present invention does not need iteration, and the calculation time only requires 20% of the AIA iteration algorithm. When the interference image quality is poor and there is a severe background light intensity fluctuation, the method of the present invention has a smaller calculation error than the AIA algorithm.

[0018] In addition, the computational effort of the method of the present invention is concentrated on solving the nonlinear equations of each pixel in the interference pattern, and the computations between the pixels in this computational process are completely independent. This feature makes the method of the present invention extremely suitable for multi-core parallel computing, thereby easily reducing the current computation time by more than 75%.

[0019] When there is an inevitable phase shift error in the optical phase shift interferometer system, the method of the present invention is particularly suitable for precise measurement of samples with small microstructures or extremely high surface processing accuracy, such as micro-nano sensors, precision optical elements, semiconductor wafers, etc. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] Figure 1 Flow chart of the optical interference phase extraction method based on principal component analysis and vector inverse orthogonalization of the present invention

[0021] Figure 2 The phase calculation results of the PCAVR algorithm, PCA+LSM algorithm and AIA algorithm of the present invention under different background noise conditions. Among them, (a) is the ideal interference image under 0% noise, (b) is the interference image under 15% background noise, (c) is the theoretical phase of the simulated inclined plane, and (d) is the residual of the PCAVR algorithm under 15% background noise.

[0022] Figure 3 The calculation consumption of the PCAVR algorithm, PCA+LSM algorithm and AIA algorithm of the present invention under different background noise conditions is shown in FIG.

[0023] Figure 4The phase calculation results of the PCAVR algorithm, PCA+LSM algorithm and AIA algorithm of the present invention under the condition of 3% background noise modulation degree change. Among them, (a) is the interference image with 3% background noise and modulation degree change, (b) is the phase residual of the PCAVR algorithm, (c) is the phase residual of the PCA+LSM algorithm, and (d) is the phase residual of the AIA algorithm.

[0024] Figure 5 The phase calculation results of the PCAVR algorithm, PCA+LSM algorithm and AIA algorithm of the present invention under the conditions of 3% background noise, background light intensity and modulation degree changes. Among them, (a) is the original interference image with 5% background noise and modulation degree changes, (b) is the phase residual of the PCAVR algorithm, (c) is the phase residual of the PCA+LSM algorithm, and (d) is the phase residual of the AIA algorithm. DETAILED DESCRIPTION

[0025] The embodiments of the present invention are described in detail below with reference to the accompanying drawings and examples.

[0026] The present invention discloses an optical interference phase extraction method based on principal component analysis and vector inverse orthogonalization. The light intensity value of the same pixel point in the interference image set is regarded as a multidimensional light intensity vector, and according to the double-beam interference light intensity formula, the light intensity vectors of different pixel points are regarded as the result of linear representation by a set of unknown initial basis vectors. The phase shift between the unknown initial basis vectors and the interference image is a strict trigonometric function relationship. By solving the unknown initial basis vectors, the real phase shift between the interference images can be accurately calculated, and finally the wrapped phase of the object surface height is solved according to the obtained interference image phase shift. When performing the calculation, a set of standard orthogonal bases of the interference light intensity vectors are first solved by the principal component analysis method, and the first two principal elements of the standard orthogonal bases are regarded as the results after the unknown initial basis vectors are orthogonalized. Subsequently, constraint conditions are set according to the trigonometric function relationship of the initial basis vectors to establish a set of equations, and the initial basis vectors are obtained by vector inverse orthogonal calculation, thereby calculating the real phase shift between the interference images. Finally, the actual phase shift between the interference images is substituted into the linear overdetermined equations constructed based on the double-beam interference intensity theory to obtain the least squares solution of the highly wrapped phase on the surface of the object.

[0027] like Figure 1 As shown, the method of the present invention mainly comprises the following steps:

[0028] Step 1: reconstruct the interference image data according to the requirements of the principal component analysis method to obtain a light intensity matrix I, wherein the number of interference images is at least 4.

[0029] The phase-shift interference image set of the measured object of the present invention is collected by an optical phase-shift interference device. The collected interference image should have good fringe contrast. On this basis, the interference pattern is reconstructed according to the requirements of the principal component analysis method to obtain the light intensity matrix I.

[0030] According to the principle of optical interference, the data reconstruction principle and specific method of this step are as follows: the gray value of each pixel in the phase-shift interference image is equal to the interference light intensity value, and the interference light intensity values ​​of the same pixel in different interference images constitute an n-dimensional interference light intensity vector I(x, y), which is expressed as follows:

[0031] I(x,y)=[I1(x,y),I2(x,y),…,I n (x,y)]

[0032] Among them I n (x, y) represents the interference light intensity value of the pixel (x, y) in the nth interference image. The light intensity matrix I is composed of the interference light intensity vectors of all pixels, which is expressed as follows:

[0033] I=[I(1,1),I(1,2),I(1,3),…,I(N x ,N y )]

[0034] The matrix has a total of n rows N x ×N y Column, n is the number of interference patterns, N x ×N y is the resolution of the interference image. Furthermore, the interference light intensity value I of the pixel point (x, y) in the nth interference image n (x,y) is represented as:

[0035] I n (x,y)=A(x,y)+B(x,y)cos[Φ(x,y)+δ n ]

[0036] A(x,y) is the background light intensity at the pixel point, B(x,y) is the modulation degree of the interference signal at the pixel point, and Φ(x,y) is the height phase of the object surface. Assuming that there are n interference images in the interference image set, the light intensity values ​​of the same pixel in different interference images are taken as a set of data, and the light intensity expression is expanded by trigonometric functions to obtain the following set of equations.

[0037]

[0038] The system of equations can be rewritten in vector form I(x,y)=α(x,y)V0+β(x,y)V1+γ(x,y)V2

[0039] Where I(x,y)=[I1(x,y),I2(x,y),…,I n (x,y)] T , V0=[1,1,…,1] T , V1=[cosδ1,cosδ2,…,cosδ n ] T , V2=[sinδ1,sinδ2…,sinδ n ] T , [T] is the transposition operation. According to the basic theory of optical interference, we know that α(x,y)=A(x,y), β(x,y)=B(x,y)cosΦ(x,y), γ(x,y)=-B(x,y)sinΦ(x,y). A(x,y) and B(x,y) represent the background light intensity and modulation degree of the interference signal at the pixel coordinate (x,y), respectively. Φ(x,y) is the surface height phase of the object to be determined at the pixel. From the above formula, we can see that the vector I(x,y) formed by different pixel points can be linearly represented by a set of initial basis vectors V0, V1, V2 that are independent of the surface morphology of the object, and the initial basis vectors V1 and V2 contain the real phase shift information between different interference images. Removing the background light intensity in the above expression can be obtained at this time It is completely linearly represented by the initial basis vectors V1 and V2. The present invention uses the principal component analysis method and the inverse orthogonalization of the principal element vector to solve the initial basis vectors V1 and V2, thereby obtaining the accurate phase shift between the interference patterns. Finally, the solved accurate phase shift is substituted into the overdetermined linear equations to obtain the true phase Φ(x, y) of the object surface.

[0040] Step 2: Calculate the covariance matrix C of the matrix I, C = (Im x )(Im x ) T , matrix m x The same size as the intensity matrix I, matrix m x All elements in are equal and are the mean values ​​of all elements in the light intensity matrix I. Perform principal component analysis to obtain the principal element vectors, and take the two principal element vectors p1 and p2 corresponding to the two largest eigenvalues ​​λ1 and λ2.

[0041] The specific implementation method of performing principal component analysis on the covariance matrix C to obtain each principal element vector is as follows:

[0042] (Im x ) can centralize the data, which can remove and suppress the background light intensity. The matrix m x All elements in are equal and are the mean of all elements of the interference intensity matrix I, while in other principal component analysis-based algorithms, the matrix m xAll elements in each row are equal and are the mean of all elements in the corresponding row of the interference intensity matrix I. This change takes into account that when the fringe period in the interference image is less than 1 or the number of fringes is small, one interference image may be darker overall, while the next interference image may be brighter overall. The background intensity calculated from a single interference image may have a large error with the actual value. Therefore, let the matrix m x All elements in are equal and are the mean of all elements of the interference light intensity matrix I to better suppress the background light intensity. Since C is a real symmetric matrix, according to the theory of linear algebra, it is certain that a transformation matrix D can be found to diagonalize C, that is, F = DCD T , where D is composed of the eigenvectors of the covariance matrix C, which are orthogonal to each other. F is the result of diagonalizing the C matrix, and the elements on its diagonal are the eigenvalues ​​of the C matrix. The singular value decomposition method and Hotelling transformation algorithm are used to complete the matrix diagonalization calculation, and the solved eigenvalues ​​are recorded from large to small as λ1, λ2, …, λ n , the eigenvectors corresponding to the eigenvalues, i.e., the principal element vectors, are denoted as p1, p2, ..., p n .

[0043] Step 3: According to the constraints of the basis vectors, the orthogonal principal element vectors p1 and p2 are deorthogonalized to solve the initial basis vectors V1 and V2.

[0044] The non-strictly orthogonal and unknown initial basis vectors V1 and V2 can be solved by inverse orthogonalization of the orthogonal principal element vectors p1 and p2 under the constraints of trigonometric functions. The specific solution method for the initial basis vectors V1 and V2 in this step: Principal component analysis is essentially to find a set of standard orthogonal bases to replace the initial basis vectors to re-represent and reduce the dimension of the vector data. The standard orthogonal bases are linearly related to the initial basis vectors. In the calculation process of the covariance matrix C, the operation (Im x ) actually suppresses and removes the influence of background light intensity, i.e., α(x,y)V0. The difference in interference light intensity between pixels is determined by the surface phase Φ(x,y) of the object and the modulation degree B(x,y) of the interference signal. As can be seen from the previous analysis, the pixel vector of the interference pattern after removing the background light intensity is It can be regarded as a linear representation of a set of fixed initial basis vectors V1, V2. And it is considered that p1 and p2 are the results of orthogonalization of the initial basis vectors V1 and V2, so the initial basis vectors V1 and V2 can be linearly represented by the first two principal components p1 and p2 solved by the principal component analysis method in step 2. According to the relationship between V1, V2 and p1, p2, the initial basis vector is represented as Among them, k1, k2, k3, and k4 are unknown coefficients to be solved, and they are solved using the least squares method. The values ​​of each dimension of the initial basis vectors V1 and V2 are the sine or cosine of the phase shift between the interference images, so the initial basis vectors satisfy the constraint V1(n) 2 +V2(n) 2 =1, where V1(n) and V2(n) represent the values ​​of the initial basis vectors V1 and V2 in the nth dimension, respectively. Therefore, there are n constraint equations in total. According to the constraint conditions, the following quadratic equation system can be obtained.

[0045]

[0046] The system of equations contains 4 unknowns, so this method requires more than four interference images to solve the quadratic equation. The following is the specific process of using the Levenberg-Marquard algorithm to solve the unknowns k1, k2, k3, k4:

[0047] 3.1、Create a vector function f(k), f(k)=[f1,f2,..,f n ] T ,k=[k1,k2,k3,k4] T And the error evaluation function F(k),

[0048]

[0049] 3.2. Given the initial iteration vector k (0) And the initial trust region radius u0, and start iterating.

[0050] u0=maxElement[J T (k (0) )J(k (0) )]

[0051] Where J(k) is the Jacobian matrix of the vector function f(k), and maxElement(x) means taking the maximum element of the matrix.

[0052]

[0053] 3.3. When iterating to the mth time, calculate the corresponding Jacobian matrix J[k (m) ] and the approximate Hessian matrix H[k (m) ]=J T [k (m) ]J[k (m) ].

[0054] 3.4 Solve the linear equations (H+μ m E) Δk=-J Tf,μ>0, and the step increment Δk is obtained, where E is the unit matrix.

[0055] 3.5. Calculate the gain ratio ρ and correct the trust region radius u for the next iteration m+1 .

[0056]

[0057] If ρ≤0.25, it means that the Taylor approximation is poor and the trust region radius should be reduced. m+1 =0.5u m .

[0058] If 0.25<ρ<0.75, it means that the Taylor approximation effect is moderate and the current trust region radius should be maintained unchanged. m+1 =u m .

[0059] If ρ ≥ 0.75, it means that the Taylor approximation is very accurate and the radius of the trust region should be expanded. Then u m+1 =2u m .

[0060] 3.6. If ρ>0, let k (m+1) =k (m) +Δk, otherwise let k (m+1) =k (m) .

[0061] 3.7. Determine whether the algorithm converges. If it does not converge or the maximum number of iterations given by the user is not reached, continue iterating. Otherwise, stop iterating and output the solution results of k1, k2, k3, and k4.

[0062] Step 4: Perform an inverse tangent operation on the initial basis vectors V1 and V2 to obtain the wrapped phase shift between the interference images. Then perform one-dimensional phase unwrapping and alignment flip operations on the wrapped phase shift between the interference images to obtain the true phase shift δ between the interference images. n .

[0063] In this step, the phase shift between the interference images is obtained by performing an inverse tangent operation and a one-dimensional phase unwrapping operation on the initial basis vectors V1 and V2. The formula is as follows:

[0064]

[0065] Where unwrap(x) means unwrapping the one-dimensional phase.

[0066] Subsequently, due to the uncertainty of the correspondence between V1 and V2, this may cause The actual phase shift δ n are inverse numbers of each other, so we need The global sign of is corrected, that is, the alignment flip operation is introduced. In the process of phase shift acquisition of interference images, the phase shift direction is determined, and it is customary to set the phase shift δ1 corresponding to the acquisition position of the first interference image to 0 rad, so it is also necessary to set Aligned to 0rad. When the phase shift direction is positive, the actual phase value δ n according to Calculation, when the phase shift direction is negative, follow Calculate, where sgn(x) is the sign function, which is defined as follows:

[0067]

[0068] in, is the misaligned phase shift corresponding to the acquisition position of the second interference image. So far, the correct phase shift between the interference images and the global sign of the phase shift have been determined.

[0069] Step 5: construct an overdetermined linear equation group containing the height phase information of the object to be measured, and obtain its least square solution to obtain the surface height phase information of the object to be measured.

[0070] The specific solution method for the surface height parcel phase information of the measured object in this step is as follows: the vector expression of the pixel light intensity in the interference pattern can be rewritten as the following overdetermined linear equation system containing the height phase information of the measured object:

[0071]

[0072] Among them, α(x,y), β(x,y), γ(x,y) are the three unknown intermediate process quantities corresponding to the pixel point (x,y). Since the equation system contains three unknown intermediate process quantities and more than four equations, its least square solution can be found. The residual function of the interference light intensity of the pixel point (x,y) is expressed as:

[0073]

[0074] The light intensity of the kth interference pattern calculated by solving the equation system, taking the partial derivative of the residual function and setting it to 0 gives the following solution:

[0075]

[0076] [α(x,y),β(x,y),γ(x,y)] T =L -1 b

[0077] in

[0078]

[0079] The above formula omits the pixel coordinates. The height phase value of each pixel is calculated by the inverse tangent trigonometric function, which is expressed as follows:

[0080]

[0081] Step six, analyzing the eigenvalues ​​obtained by principal component analysis to determine whether the phase calculation result is reliable. If the calculation result is reliable, the surface height phase information of the measured object is output and the calculation is terminated.

[0082] The reliability judgment principle and method of the phase calculation result in this step: When the background light intensity can be effectively suppressed by the mean, the light intensity vector of the background light intensity is removed. It can be completely represented by the initial basis vectors V1, V2. For principal component analysis, λ2>>λ3, or λ3≈λ4≈…≈λ n Therefore, the reliable basis for judging the phase calculation result is:

[0083]

[0084] or,

[0085]

[0086] When the above conditions are met, the phase calculation result is considered reliable, and the phase value calculated in step 5 can be directly output as the final result, where ε1 and ε2 are thresholds set by the user. The larger ε1 and the smaller ε2, the stricter the calculation reliability judgment and the smaller the final calculation error. In general, ε1 and ε2 are given as 3000 and 1.1.

[0087] Step 7: When the calculation result is judged to be unreliable, the light intensity matrix I is corrected using the solution in step 5, and steps 2 to 6 are repeated for iterative calculation until the PCAVR algorithm converges or reaches the set maximum number of iterations.

[0088] The specific correction method of the intensity vector matrix in this step is as follows: x ) effectively eliminates the background light intensity, the interference light intensity matrix is ​​corrected by α(x,y) solved in step 5 to more accurately remove the background light intensity. When the light intensity matrix I is iterated for the mth time, its correction method is:

[0089] I (m) =I (1) -G (m-1)

[0090]

[0091] Among them I (m)is the correction result at the mth iteration, that is, the light intensity vector matrix participating in the principal component analysis at the mth iteration. (1) is the intensity matrix obtained directly from the original interference image, α (m-1) (x, y) is the solution result of the unknown intermediate process quantity α(x, y) at the m-1th iteration, which is calculated by step five.

[0092] The convergence judgment basis of the PCAVR algorithm of the present invention is: Where ε3 is the convergence threshold set by the user, which can be given as 0.003. is the phase shift vector of the i-th interference pattern calculated at the m-th iteration. The symbol |·| represents the absolute value.

[0093] In a specific embodiment, a set of simulated interference images was generated using Matlab software, and the proposed phase extraction method based on principal component analysis and vector inverse orthogonalization (PCAVR) was compared with the principal component analysis and least squares fusion algorithm (PCA+LSM) and the advanced iterative algorithm (AIA) to illustrate the superiority of the proposed PCAVR algorithm in robustness and solution accuracy.

[0094] The resolution of the simulated interference image is 1000pixel×1000pixel, the actual spatial size of the pixel point is 0.345nm×0.345nm, the measured surface is an ideal inclined plane, the z-height of the inclined plane is z(x,y)=0.345xtan0.1, in nm, and the wavelength of the interference light source is 550nm. The interference image set contains a total of 5 interference images. In order to simulate the phase shift error, a random phase shift error of ±20nm is added on the basis of a single phase shift of 50nm. The theoretical random phase shift between the simulated interference images is 0rad, 1.5133rad, 1.9439rad, 3.8050rad, 4.6906rad, and they are unequal phase shifts. At the same time, in order to facilitate the display and RMS calculation of the measured results, the least squares unwrapping algorithm based on discrete cosine transform is used to unwrap the wrapped phase obtained by the PCAVR algorithm, PCA+LSM algorithm and AIA algorithm. In addition, the convergence threshold of the AIA algorithm is set to 0.0005, the thresholds ε1, ε2, and ε3 of the PCAVR algorithm are set to 3000, 1.10, and 0.003, respectively, and the maximum number of iterations is set to 12.

[0095] First, we add different degrees of background noise to the ideal interferogram and compare the performance of the above four algorithms in phase calculation. The added background noise is Gaussian noise, and the standard deviation of the noise is determined by multiplying the standard deviation of the light intensity of the interference image set by a percentage. The pixel background intensity A(x,y) in the ideal interferogram is 120, and the modulation degree B(x,y) is 85. Table 1 summarizes the calculation results of the three algorithms. Figure 2 (a) is the ideal interference image with 0% noise. Figure 2 (b) is the interference image under 15% background noise. Figure 2 (c) is the theoretical phase of the simulated slope. Figure 2 (d) is the residual of the PCAVR algorithm under 15% background noise. Figure 3 The calculation time of the three algorithms is shown. The results show that the PCA+LSM algorithm has the largest calculation error under the condition of adding only noise, and even under the condition of 0% noise, there is still a calculation error of 0.0945rad in the phase shift. Under the ideal 0 noise condition, the PCAVR algorithm can almost restore the true phase shift between the interference images and the measured inclined surface with zero error. Under various noise levels, the calculation error of the phase shift between the interference images by the PCAVR algorithm is better than 10 -4 rad level, which is one order of magnitude smaller than the AIA algorithm. When there is no background light intensity fluctuation in the interference image, the PCAVR algorithm can obtain accurate phase calculation results without iteration, while the AIA algorithm requires an average of 6 iterations to obtain the phase calculation results. In this case, the PCAVR algorithm takes only 25% of the time of the AIA algorithm. Both the PCAVR algorithm and the AIA algorithm are relatively accurate in calculating the phase shift between interference patterns. Gaussian noise is the main cause of surface calculation errors, which makes the two algorithms the same in RMS performance and both are better than the PCA+LSM algorithm.

[0096] Table 1 Calculation results of each algorithm under different noise levels

[0097]

[0098] In order to illustrate the PCAVR algorithm and its excellent performance in dealing with the fluctuation of modulation degree between pixels in the interference image, nonlinear modulation degree changes were added on the basis of adding 3% background noise (signal-to-noise ratio is 30Db). The modulation degree of the central pixel of the interference pattern is 85, and the modulation degree at the edge is 31.3. This fluctuation is huge. The modulation degree of the pixel is calculated according to the following formula.

[0099]

[0100] Table 2 shows the calculation results of the three algorithms under the conditions of 3% background noise and modulation degree change. Figure 4(a) shows the interference image with 3% background noise and modulation changes. Figure 4 (b) is the phase residual of the PCAVR algorithm. Figure 4 (c) is the phase residual of the PCA+LSM algorithm. Figure 4 (d) is the phase residual of the AIA algorithm. The calculation error and RMS of the phase shift between the interferograms of the PCAVR algorithm are 1.28×10 -4 rad and 0.0148, which are both smaller than 0.0225rad and 0.0156 calculated by AIA algorithm. The results show that PCAVR algorithm is superior to AIA algorithm in terms of both calculation speed and calculation error when the background modulation changes. PCAVR algorithm is robust to the change of modulation between pixels.

[0101] Table 2 Calculation results of various algorithms under 3% noise and modulation change conditions

[0102] Algorithm Name PCAVR PCAVR+LSM AIA <![CDATA[Errorδ n (rad)]]> <![CDATA[1.28×10 -4 ]]> 0.0790 0.0225 RMS(rad) 0.0148 0.0258 0.0156 Time(s) 5.49 4.98 23.71

[0103] Finally, nonlinear background intensity change and modulation change are added on the basis of the original 3% background noise (signal-to-noise ratio is 30dB). The background intensity change and modulation change of the pixel point in the interference pattern are changed according to the following formulas:

[0104]

[0105] The maximum and minimum values ​​of the background light intensity in the interference pattern are 120 and 44.1 respectively, and the maximum and minimum values ​​of the modulation index are 85 and 31.3 respectively. Table 3 shows the phase calculation results of the three algorithms under the conditions of adding 5% background noise and changing background light intensity and modulation index. Figure 5 (a) shows the simulated interference image. Figure 5 (b) is the phase residual using the PCAVR algorithm. Figure 5 (c) is the phase residual using the PCA+LSM algorithm. Figure 5 (d) is the phase residual using the AIA algorithm. The results show that when the background light intensity in the interferogram fluctuates, the phase shift errors of the AIA algorithm and the PCA+LSM algorithm are 0.1296rad and 0.1439rad, respectively, and the RSMs are 0.0212 and 0.0219, respectively. The PCAVR algorithm can still accurately calculate the phase shift of the interferogram through iteration, and the calculation error and RMS of the phase shift between the interferograms are 0.0065rad and 0.0161, respectively.

[0106] Table 3 Calculation results of each algorithm under the conditions of 5% noise, background light intensity and modulation degree change

[0107]

[0108]

[0109] The simulation results show that the method of the present invention can more accurately calculate the real phase shift between interference images, accurately extract the phase information of the interference image collected with any phase shift step, and is insensitive to the fluctuation of the modulation degree of the pixel points in the interference image. When there are background light intensity fluctuations and modulation degree fluctuations in the pixel points in the interference image, the calculation accuracy of the PCAVR algorithm is significantly better than that of the AIA algorithm and the PCA+LSM algorithm, which makes up for the shortcomings of the AIA algorithm. Even if there are drastic background light intensity fluctuations in the interference image, the PCAVR algorithm can accurately calculate the phase shift of the interference image and the height phase of the object being measured.

[0110] When there is an inevitable phase shift error in the optical phase shift interferometer device, the method of the present invention is particularly suitable for precise measurement of optical elements, semiconductor wafers, micro-nano sensors and other samples with small microstructures or extremely high processing accuracy. In the future, we plan to apply this algorithm to defect detection, surface shape measurement and roughness measurement of wafer surfaces.

Claims

1. An optical interference phase extraction method based on principal component analysis and vector inverse orthogonalization, characterized in that: The steps include: Step 1, reconstructing the interference image data according to the requirements of the principal component analysis method to obtain a light intensity matrix I, and the number of interference images is at least 4; Step 2: Calculate the covariance matrix C of the light intensity matrix I, and perform principal component analysis to obtain the two principal element vectors p1 and p2 corresponding to the two largest eigenvalues ​​λ1 and λ2; Step 3: According to the constraint conditions of the basis vectors, the orthogonal principal element vectors p1 and p2 are vector-inverse orthogonalized to solve the initial basis vectors V1 and V2; Step 4: Perform an inverse tangent operation on the initial basis vectors V1 and V2 to obtain the wrapped phase shift between the interference images; perform one-dimensional phase unwrapping and flip alignment on the wrapped phase shift between the interference images to obtain the true phase shift δ between the interference images n ; Step 5, constructing an overdetermined linear equation system containing the height phase information of the object to be measured, and finding its least square solution to obtain the surface height phase information of the object to be measured; Step 6, analyzing the eigenvalues ​​obtained by principal component analysis to determine whether the phase calculation result is reliable. If the calculation result is reliable, the surface height phase information of the measured object is output and the calculation is terminated; Step 7: When the calculation result is judged to be unreliable, the light intensity matrix I is corrected using the solution in step 5, and steps 2 to 6 are repeated for iterative calculation until the algorithm converges or reaches the set maximum number of iterations.

2. The optical interference phase extraction method based on principal component analysis and vector inverse orthogonalization according to claim 1, characterized in that: In the step 1, a phase-shift interference image set of the object to be measured is collected by an optical phase-shift interference device. The gray value of each pixel in the phase-shift interference image is equal to the interference light intensity value. The interference light intensity values ​​of the same pixel in different interference images constitute an n-dimensional interference light intensity vector I (x, y), which is represented as follows: I(x,y)=[I1(x,y),I2(x,y),…,I n (x,y)] Among them I n (x, y) represents the interference light intensity value of the pixel (x, y) in the nth interference image. The light intensity matrix I is composed of the interference light intensity vectors of all pixels, which is expressed as follows: I=[I(1,1),I(1,2),I(1,3),…,I(N x ,N y )] The matrix has a total of n rows N x ×N y Column, n is the number of interference patterns, N x ×N y is the resolution of the interference image.

3. The optical interference phase extraction method based on principal component analysis and vector inverse orthogonalization according to claim 2 is characterized in that: In step 2, the covariance matrix C is expressed as follows: C=(I-m x )(I-m x ) T Among them, the matrix m x The same size as the intensity matrix I, matrix m x All elements in are equal and are the mean of all elements of the light intensity matrix I; Perform principal component analysis on the covariance matrix C to obtain the principal element vectors. The implementation method is as follows: The singular value decomposition method and Hotelling transformation algorithm are used to complete the matrix diagonalization calculation, and the solved eigenvalues ​​are recorded as λ1, λ2, …, λ n , the eigenvectors corresponding to the eigenvalues, i.e., the principal element vectors, are denoted as p1, p2, ..., p n .

4. The optical interference phase extraction method based on principal component analysis and vector inverse orthogonalization according to claim 3 is characterized in that: In step 3, the constraints for solving the initial basis vectors V1 and V2 are: V1(n) 2 +V2(n) 2 =1 Among them, V1(n) and V2(n) represent the values ​​of the initial basis vectors V1 and V2 in the nth dimension respectively. The relationship between V1, V2 and p1, p2 is as follows: The unknown coefficients k1, k2, k3, and k4 are solved using the least squares method.

5. The optical interference phase extraction method based on principal component analysis and vector inverse orthogonalization according to claim 4, characterized in that: In step 4, the initial basis vectors V1 and V2 are subjected to an inverse tangent operation and a one-dimensional phase unwrapping operation to obtain the phase shift of the unaligned interference image. The formula is as follows: Here, unwrap(x) means unwrapping the one-dimensional phase.

6. The optical interference phase extraction method based on principal component analysis and vector inverse orthogonalization according to claim 5, characterized in that: The unaligned phase shift obtained by the one-dimensional phase unwrapping operation is subjected to an alignment flip operation to obtain the final phase shift. The method is: Let the actual phase shift δ1 corresponding to the acquisition position of the first interference image be 0rad, then Align to 0rad; When the phase shift direction is positive: When the phase shift direction is negative: Where sgn(x) is the sign function, which is defined as follows is the misaligned phase shift corresponding to the acquisition position of the second interference image.

7. The optical interference phase extraction method based on principal component analysis and vector inverse orthogonalization according to claim 3, characterized in that: The step 5 is to construct an overdetermined linear equation system containing the height phase information of the object being measured, which is expressed as follows: Among them, α(x,y), β(x,y), and γ(x,y) are three unknown intermediate process quantities corresponding to the pixel point (x,y). The residual function of the interference light intensity of the pixel point (x,y) is expressed as: Taking the partial derivative of the residual function and setting it to 0 gives the following solution: [α(x,y),β(x,y),γ(x,y)] T =L -1 b in The height phase value of each pixel is calculated by the inverse tangent trigonometric function, which is expressed as follows:

8. The optical interference phase extraction method based on principal component analysis and vector inverse orthogonalization according to claim 3, characterized in that: In step 6, the basis for judging the reliability of the phase calculation result is: or, Among them, ε1 and ε2 are the set thresholds.

9. The optical interference phase extraction method based on principal component analysis and vector inverse orthogonalization according to claim 8, characterized in that: In step 7, the correction method of the light intensity matrix I during the mth iteration is: I (m) =I (1) -G (m-1) Among them I (m) is the modified result at the mth iteration, I (1) is the light intensity matrix obtained directly from the interference pattern, α (m-1) (x, y) is the solution result of the unknown intermediate process quantity α(x, y) at the m-1th iteration, which is calculated by step five.

10. The optical interference phase extraction method based on principal component analysis and vector inverse orthogonalization according to claim 9, characterized in that: In step 7, the algorithm convergence judgment basis is Where ε3 is the set convergence threshold, is the phase shift vector of the i-th interference pattern calculated at the m-th iteration. The symbol |·| represents the absolute value.

Citation Information

Patent Citations

  • Principal component analysis phase retrieval algorithm

    CN103778650A

  • Method for recovering phase distribution of phase shift interference figures and method for obtaining phase shift between two figures

    CN104614083A

  • Vibration measurement system and method based on multi-longitudinal-mode orthogonal laser self-mixing sensor

    CN117330172A

  • Method and device for analysing an electromagnetic wave in high definition

    IN201917015305A

  • Beamforming method, measurement imaging device, and communication device

    JP2019030623A