In-situ measurement method and system for wavefront aberration of objective lens of optical quantity detection device

Abstract

The present disclosure provides an objective wave aberration in-situ measurement method and system of an optical quantity detection device. The construction method comprises: establishing a vector imaging forward model based on the hardware parameters of the optical quantity detection device to be measured; generating a plurality of sets of Zernike aberration coefficient samples within a preset aberration fluctuation range, and calculating a simulated defocus image set of a plurality of detection marks under each Zernike aberration coefficient sample and a plurality of defocus positions by using the vector imaging forward model; performing data dimension reduction processing on the simulated defocus image set corresponding to each detection mark to obtain a plurality of target feature components and target feature coefficients corresponding to each detection mark; and constructing a mapping relationship model based on the target feature coefficients and the Zernike aberration coefficient samples. The present disclosure can realize comprehensive and efficient sampling of the pupil.

Description

Technical Field

[0001] This disclosure relates to the field of in-situ measurement technology of objective wavelet aberration of optical measurement equipment, and in particular to a method and system for in-situ measurement of objective wavelet aberration of optical measurement equipment. Background Technology

[0002] In the semiconductor manufacturing and inspection process, high-resolution optical measurement equipment, such as defect detection equipment and overlay error measurement equipment, is the core equipment for realizing wafer and mask defect detection and critical dimension measurement, and plays an important role in monitoring process stability and ensuring product yield.

[0003] The imaging quality of the objective lens directly affects the measurement accuracy and reliability of optical measurement equipment, and objective lens wavefront aberration is one of the key factors determining image quality. Objective lens wavefront aberration can cause image plane distortion, decreased contrast, and degradation of resolving power. At the same time, during long-term, high-load operation, the objective lens wavefront aberration will dynamically change due to factors such as illumination thermal effects, mechanical stress, and environmental drift, which can lead to missed detection of minute defects, false detection of normal structures, and deviations in measurement results.

[0004] Existing in-situ methods for objective wavelet aberration measurement typically employ a single type of detection marker combined with multiple illumination modes. For example, they might use periodic gratings in different directions or a single two-dimensional detection marker to acquire defocused images, followed by aberration inversion. However, these methods have significant shortcomings in practical applications: First, the spatial frequency components excited by a single type of detection marker are limited, resulting in insufficient coverage of the pupil plane, low pupil sampling rate, and insufficient sensitivity to certain types of aberrations. Second, different aberration terms exhibit varying response characteristics to the detection marker, making it difficult for a single type of marker to simultaneously meet the measurement needs of multiple aberration terms, especially higher-order Zernike aberration terms, thus limiting the number of accurately measurable aberration terms. Third, to compensate for the insufficient information from a single type of marker, existing methods often require frequent switching of illumination modes, leading to complex measurement procedures, long image acquisition times, and low marker image sampling efficiency, failing to meet the demands of modern optical measurement equipment for efficient, comprehensive, and high-precision in-situ aberration measurement. Summary of the Invention

[0005] The technical problem to be solved by this disclosure is to overcome the shortcomings of the prior art in that it is impossible to perform comprehensive, efficient and high-precision in-situ aberration measurement of optical quantity detection equipment, and to provide a method and system for in-situ measurement of objective wavelet aberration of optical quantity detection equipment.

[0006] This disclosure solves the above-mentioned technical problems through the following technical solution:

[0007] A first aspect of this disclosure provides a method for constructing a mapping relationship model for in-situ measurement of objective wavelet aberrations, the method comprising:

[0008] A vector imaging forward model is established based on the hardware parameters of the optical quantity detection device to be measured.

[0009] Multiple sets of Zernike aberration coefficient samples are generated within a preset aberration fluctuation range. The vector imaging forward model is used to calculate the simulated defocus image set of various detection markers at each of the Zernike aberration coefficient samples and multiple defocus positions. The spatial frequency response features of the various detection markers are complementary.

[0010] Data dimensionality reduction processing is performed on the simulated defocused image set corresponding to each detection marker to obtain several target feature components and their target feature coefficients corresponding to each detection marker.

[0011] A mapping relationship model is constructed based on the target feature coefficients and the Zernike aberration coefficient samples.

[0012] Optionally, the step of constructing a mapping relationship model based on the target feature coefficients and the Zernike aberration coefficient samples includes:

[0013] Based on the target feature coefficients, the aberration sensitivity of each detection label to each Zernike aberration term is evaluated, and the target feature components of each detection label are fused based on the aberration sensitivity evaluation results to construct a multi-detection label joint feature vector corresponding to each Zernike aberration term.

[0014] The mapping relationship model is constructed based on the target feature coefficients, the joint feature vector of the multi-detection labels, and the Zernike aberration coefficient samples.

[0015] Optionally, the step of evaluating the aberration sensitivity of each detection marker to each Zernike aberration term based on the target feature coefficients includes:

[0016] Obtain the first feature coefficient of the same detection marker under the first Zernike aberration coefficient sample and the second feature coefficient under the second Zernike aberration coefficient sample;

[0017] The target Zernike aberration coefficients of the first Zernike aberration coefficient sample and the second Zernike aberration coefficient sample are different.

[0018] Calculate the first variation magnitude of the first feature coefficient and the second feature coefficient, and the second variation magnitude of the target Zernike aberration term coefficient of the first Zernike aberration coefficient sample and the second Zernike aberration coefficient sample;

[0019] The aberration sensitivity of the detection marker to the target Zernike aberration term is determined based on the first change magnitude and the second change magnitude.

[0020] Optionally, the step of performing data dimensionality reduction processing on the simulated defocus image set corresponding to each detection marker to obtain several target feature components and their target feature coefficients corresponding to each detection marker includes:

[0021] Perform dimensionality reduction processing on the simulated defocused image set corresponding to each detection marker to obtain the candidate feature components and their candidate feature coefficients corresponding to each detection marker.

[0022] The third aberration sensitivity of each candidate feature component to each Zernike aberration term is determined based on the candidate feature coefficients.

[0023] The candidate feature components whose third aberration sensitivity is greater than a preset threshold are selected to obtain the target feature components and their target feature coefficients;

[0024] And / or,

[0025] The step of constructing a mapping relationship model based on the target feature coefficients and the Zernike aberration coefficient samples includes:

[0026] The target feature coefficients and the Zernike aberration coefficient samples are fitted using a fitting algorithm to obtain the mapping relationship model;

[0027] And / or,

[0028] The spatial frequency range of the various detection markers under a single illumination mode fully covers the effective transmission frequency range of the pupil;

[0029] And / or,

[0030] The detection markings include at least two of the following: circular hole markings, circular ring markings, and square ring markings;

[0031] And / or,

[0032] The step of performing data dimensionality reduction processing on the simulated off-focus image set corresponding to each of the detection markers includes:

[0033] Principal component analysis and / or singular value decomposition are used to extract features and reduce dimensions of the simulated defocused image set corresponding to each detection marker.

[0034] And / or,

[0035] The mapping relationship model is established using at least one of multiple linear regression, partial least squares regression, and artificial neural networks.

[0036] A second aspect of this disclosure provides a method for in-situ measurement of wavelet aberration of an objective lens in an optical measurement device, the method comprising:

[0037] Multiple detection marks are set up within the same field of view of the optical quantity detection equipment to be measured;

[0038] The device for detecting the optical quantity under test is driven to acquire a set of actual multi-marker defocused images at multiple defocused positions;

[0039] The actual multi-marker defocused image set is subjected to data dimensionality reduction processing to obtain the feature components of each target and their actual feature coefficients;

[0040] The actual characteristic coefficients are input into the mapping relationship model to obtain the actual Zernike aberration coefficients of the objective lens in the optical quantity detection device to be measured;

[0041] The mapping relationship model is pre-constructed based on the construction method described in the first aspect of this disclosure.

[0042] Optionally, before the step of performing data dimensionality reduction processing on the actual multi-labeled off-focus image set, the method further includes:

[0043] The actual multi-marker defocus image set is preprocessed;

[0044] The preprocessing includes at least one of normalization, background correction, noise suppression, and image registration;

[0045] And / or,

[0046] The actual multi-marker defocused image set was obtained under a single illumination mode;

[0047] And / or,

[0048] The optical quantity detection equipment includes one of the following: a transmission-reflection optical quantity detection equipment, a total reflection optical quantity detection equipment, and a reflection optical quantity detection equipment that includes a vertical incident light path and an oblique incident light path.

[0049] A third aspect of this disclosure provides a system for constructing a mapping relationship model for in-situ measurement of objective wavelet aberrations, the system comprising:

[0050] A module is established to build a forward vector imaging model based on the hardware parameters of the optical quantity to be measured detection device.

[0051] The calculation module is used to generate multiple sets of Zernike aberration coefficient samples within a preset aberration fluctuation range, and to use the vector imaging forward model to calculate the simulated defocus image sets of various detection markers at each of the Zernike aberration coefficient samples and multiple defocus positions; wherein, the spatial frequency response features of the various detection markers are complementary.

[0052] The processing module is used to perform data dimensionality reduction processing on the simulated defocus image set corresponding to each detection mark, and obtain several target feature components and their target feature coefficients corresponding to each detection mark;

[0053] The module is used to construct a mapping relationship model based on the target feature coefficients and the Zernike aberration coefficient samples.

[0054] A fourth aspect of this disclosure provides an in-situ measurement system for objective wavelet aberration of an optical quantity detection device, the system comprising:

[0055] The acquisition module is used to deploy multiple detection marks within the same field of view of the optical quantity detection device under test, and drive the optical quantity detection device under test to acquire a set of actual multi-marker defocus images at multiple defocus positions;

[0056] The acquisition module is used to perform data dimensionality reduction processing on the actual multi-marker defocus image set to obtain each target feature component and its actual feature coefficients;

[0057] The measurement module is used to input the actual characteristic coefficients into the mapping relationship model to obtain the actual Zernike aberration coefficients of the objective lens in the optical quantity detection device to be measured.

[0058] The mapping relationship model is pre-built based on the construction system described in the third aspect of this disclosure.

[0059] A fifth aspect of this disclosure provides an electronic device including a memory, a processor, and a computer program stored in the memory and for running on the processor. When the processor executes the computer program, it implements the method for constructing a mapping relationship model for in-situ measurement of objective wavelet aberrations as described in the first aspect of this disclosure, or the method for in-situ measurement of objective wavelet aberrations of an optical quantity detection device as described in the second aspect of this disclosure.

[0060] A sixth aspect of this disclosure provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the method for constructing a mapping relationship model for in-situ measurement of objective wavelet aberrations as described in the first aspect of this disclosure, or the method for in-situ measurement of objective wavelet aberrations of an optical quantity detection device as described in the second aspect of this disclosure.

[0061] A seventh aspect of this disclosure provides a computer program product, including a computer program that, when executed by a processor, implements the method for constructing a mapping relationship model for in-situ measurement of objective wavelet aberrations as described in the first aspect of this disclosure, or the method for in-situ measurement of objective wavelet aberrations of an optical quantity detection device as described in the second aspect of this disclosure.

[0062] Based on common knowledge in the field, the above optional conditions can be combined arbitrarily to obtain the optional examples of this disclosure.

[0063] The positive advancements of this disclosure are as follows: by combining multiple detection markers with complementary spatial frequency response characteristics, comprehensive and efficient sampling of the pupil is achieved, significantly improving the measurability and measurement accuracy of more Zernike aberration terms, especially higher-order aberration terms. Furthermore, it enables targeted optimization design of detection markers for specific Zernike aberration types, allowing different detection markers to maintain high response sensitivity to different aberration terms, thereby overcoming the limitation of a single detection marker being insufficient to cover multiple types of aberrations. At the same time, by performing dimensionality reduction decomposition and sensitivity screening on the feature coefficients, redundant information and low-sensitivity features are effectively eliminated, enhancing the robustness of the aberration inversion process, improving computational efficiency, and reducing the impact of noise and disturbances on the measurement results. Furthermore, it can simultaneously acquire images of multiple detection marks within the same field of view under a single illumination mode, reducing illumination mode switching and repeated shooting processes, optimizing the image measurement process, improving the sampling efficiency of the mark image and the overall measurement efficiency, and eliminating the need for additional hardware modifications to existing optical quantity detection equipment. It can be implemented directly using the existing imaging system and area array detector, has strong compatibility, is easy to integrate into engineering, and can achieve non-invasive, high-precision in-situ measurement of objective wavelet aberrations. Attached Figure Description

[0064] Figure 1 This is a flowchart illustrating the method for constructing the mapping relationship model disclosed herein;

[0065] Figure 2 This is a schematic diagram of the transmission-reflection optical quantity detection device disclosed herein.

[0066] Figure 3 This is a schematic diagram of the total internal reflection optical quantity detection device disclosed herein;

[0067] Figure 4 This is a schematic diagram of the structure of the reflective optical quantity detection device disclosed herein;

[0068] Figure 5 This is a schematic diagram of the three detection markers disclosed herein;

[0069] Figure 6 This is a schematic diagram showing the comparison of multi-order Zernike wave aberration coefficients in this disclosure;

[0070] Figure 7 This is a schematic diagram showing the comparison of absolute errors of the multi-order Zernike wave aberration coefficients disclosed in this paper.

[0071] Figure 8 This is a flowchart of the in-situ measurement method for objective wavelet aberrations disclosed herein;

[0072] Figure 9 A flowchart serving as a specific example of this disclosure;

[0073] Figure 10 This is a schematic diagram of the modules of the system for constructing the mapping relationship model of this disclosure;

[0074] Figure 11 This is a schematic diagram of the modules of the objective wavelet aberration in-situ measurement system disclosed herein;

[0075] Figure 12 This is a schematic diagram of the structure of the electronic device disclosed herein. Detailed Implementation

[0076] The present disclosure is further illustrated below by way of embodiments, but the present disclosure is not limited to the scope of the embodiments described herein.

[0077] The prefixes such as "first" and "second" used in this disclosure are merely for distinguishing different descriptive objects and do not limit the position, order, priority, quantity, or content of the described objects. The use of ordinal numbers and other prefixes used to distinguish descriptive objects in this disclosure does not constitute a limitation on the described objects. The description of the described objects is given in the claims or the context of the embodiments, and should not be construed as an unnecessary limitation. Furthermore, in the description of this embodiment, unless otherwise stated, "multiple" means two or more.

[0078] Example 1

[0079] In one specific embodiment of this disclosure, a method for constructing a mapping relationship model for in-situ measurement of objective wavelet aberrations is provided, such as... Figure 1 As shown, the construction method includes:

[0080] S11. Establish a vector imaging forward model based on the hardware parameters of the optical quantity detection device to be measured;

[0081] S12. Generate multiple sets of Zernike aberration coefficient samples within a preset aberration fluctuation range, and use the vector imaging forward model to calculate the simulated defocus image sets of various detection markers at each Zernike aberration coefficient sample and multiple defocus positions; wherein, the spatial frequency response characteristics of various detection markers are complementary.

[0082] S13. Perform data dimensionality reduction processing on the simulated defocused image set corresponding to each detection mark to obtain several target feature components and their target feature coefficients corresponding to each detection mark.

[0083] S14. Construct a mapping relationship model based on the target feature coefficients and Zernike aberration coefficient samples.

[0084] Specifically, step S11 establishes a vector imaging forward model based on the actual hardware parameters of the optical quantity detection device under test. These actual hardware parameters may include the illumination source wavelength, illumination mode, objective lens numerical aperture (NA), polarization state distribution, imaging magnification, pixel size of the area array detector, and noise characteristics. The optical quantity detection device under test may be one of the following: a transmission-reflection optical quantity detection device, a total internal reflection optical quantity detection device, or a reflective optical quantity detection device that includes both perpendicular and oblique incident light paths.

[0085] like Figure 2 As shown, the transmission-reflection optical quantity detection equipment mainly includes an illumination source 11, an illumination optical path 12, a first illumination beam shaping lens group 12a, a second illumination beam shaping lens group 12b, a workpiece stage 13 carrying the test target, a test objective lens 14, a first objective lens imaging lens 14a, a second objective lens imaging lens 14b, a beam splitter 16, and a detector 15. Figure 3 As shown, the total internal reflection optical quantity detection equipment mainly includes an illumination source 21, a reflective illumination optical path 22, a first illumination beam shaping reflector group 22a, a second illumination beam shaping reflector group 22b, a workpiece stage 23 carrying the test target, a reflective objective lens under test 24, a first objective lens imaging reflector group 24a, a first objective lens imaging reflector group 24b, a first objective lens imaging reflector group 24c, and a detector 25. Figure 4 As shown, the reflective optical quantity detection device, which includes a vertical incident light path and an oblique incident light path, mainly includes illumination sources 31 and 36, a vertical illumination light path 32, a first vertical incident illumination beam shaping lens group 32a, a second vertical incident illumination beam shaping lens group 32b, an oblique incident illumination light path 37, a first oblique incident illumination beam shaping lens group 37a, a first oblique incident illumination beam shaping lens group 37b, a workpiece stage 33 carrying a test target, a vertical light path objective lens 34, a first vertical incident objective lens imaging lens group 34a, a second vertical incident objective lens imaging lens group 34b, an oblique incident light path objective lens 38, a first oblique incident objective lens imaging lens group 38a, a first oblique incident objective lens imaging lens group 38b, a beam splitter 40, and detectors 35 and 39.

[0086] For example, when the optical quantity to be measured is detected by a deep ultraviolet (DUV) defect detection device, the illumination wavelength used is 193nm (nanometer) or 266nm, the illumination mode can be conventional illumination or ring illumination, and the detector can be a CCD (charge-coupled device) or CMOS (complementary metal-oxide-semiconductor) array detector, then the corresponding vector imaging forward model can be established according to the illumination wavelength, illumination mode and detector parameters.

[0087] After establishing the vector imaging forward model, based on the imaging response characteristics corresponding to different Zernike aberration terms and the spatial frequency response characteristics of different detection markers, the detection markers are selectively chosen and optimized. K types of detection markers with complementary spatial frequency response characteristics are selected, ensuring that each marker maintains high sensitivity to different aberration terms. This allows the spatial frequency range of multiple detection markers under a single illumination mode to comprehensively cover the effective transmission frequency range of the pupil plane, improving the measurement capability for more orders of Zernike aberration terms, especially higher-order aberration terms. Here, K is a positive integer greater than 1, and the size of the K detection markers can be determined according to the resolution. For example, the size of the detection markers is determined based on the minimum resolution. The K detection markers are generally evenly arranged across the entire field of view, such as in 2x2, 3x3, or 1x5 arrays.

[0088] For example, such as Figure 5 As shown, the spatial frequency distribution and diffraction order characteristics excited on the pupil surface by the three detection markers—"circular aperture," "circular ring," and "square ring"—exhibit significant differences and exhibit good complementarity. Therefore, these three detection markers can be selected as measurement markers. Furthermore, different detection markers have varying sensitivities to different aberration terms. Therefore, using multiple detection markers in combination can effectively overcome the shortcomings of a single marker in terms of pupil sampling rate, aberration sensitivity, and information completeness, achieving comprehensive and efficient sampling of the pupil.

[0089] Of course, the detection mark can also be any combination of any two of the three types: "circular hole", "circular ring" and "square ring", or it can be any one or two of the above detection marks combined with one or more other shaped detection marks. This specific embodiment does not limit this.

[0090] Step S12, based on the determined M Zernike aberration term base (e.g., 33 Zernike aberration terms) within a preset aberration fluctuation range, generates multiple sets of Zernike aberration coefficient samples matching the M Zernike aberration term base using a random mixing method. Multiple sets of wavelet aberration parameters are sampled, and combined with defocus scanning parameters, a simulated defocus image set for each detection marker at different defocus positions is calculated using a vector imaging forward model. Here, M is a positive integer greater than 1.

[0091] For example, the 5th to 37th terms of the first 37 Zernike polynomials are selected as the aberration terms to be inverted, and a total of M=33 Zernike aberration coefficients are used as input variables. Each aberration coefficient can be randomly generated in the range of [-0.03λ, 0.03λ], and multiple sets of Zernike aberration coefficient samples can be generated. The defocus scanning range can be centered on the focal plane and offset by 1000nm in each positive and negative direction along the Z-axis, with a step value of 100nm. The forward vector imaging model is used to simulate imaging for different combinations of Zernike aberration coefficients, different defocus positions, and different detection marks, and the simulated defocus image set corresponding to each detection mark is obtained.

[0092] Step S13 performs dimensionality reduction processing on the obtained simulated defocus image sets for each detection marker, such as feature extraction and dimensionality reduction decomposition, to extract low-dimensional features characterizing the change patterns of the defocused images. By reducing the dimensionality of the image data for different detection markers, N target feature components and their target feature coefficients corresponding to each detection marker can be obtained, where N is a positive integer greater than 1, and the target feature components are different for different detection markers. Principal Component Analysis (PCA), Singular Value Decomposition (SVD), or other algorithms capable of high-dimensional image feature compression can be used to perform feature extraction and dimensionality reduction decomposition on the image data.

[0093] Step S14 involves performing M-term Zernike aberration basis fusion on each detection marker's target feature components and their target feature coefficients to obtain a (KN)×M-dimensional target feature coefficient matrix. The target feature coefficient matrix is ​​then fitted with multiple sets of 1×M-dimensional Zernike aberration coefficient samples to construct a mapping model, which is used in subsequent actual measurement processes to quickly invert the objective wavelet aberration coefficients from the image feature vectors.

[0094] This specific implementation combines multiple detection markers with complementary spatial frequency response characteristics to achieve comprehensive and efficient sampling of the pupil, significantly improving the measurability and measurement accuracy of more Zernike aberration terms, especially higher-order aberration terms, thereby overcoming the limitation of a single detection marker being insufficient to cover multiple types of aberrations.

[0095] In one specific embodiment, step S14 includes:

[0096] S141. Based on the target feature coefficients, evaluate the aberration sensitivity of each detection label to each Zernike aberration term, and based on the aberration sensitivity evaluation results, fuse the target feature components of each detection label to construct a joint feature vector of multiple detection labels corresponding to each Zernike aberration term.

[0097] S142. Based on the target feature coefficients, the joint feature vector of the multi-detection labels, and the Zernike aberration coefficient samples, construct the mapping relationship model.

[0098] Specifically, in order to perform targeted optimization design of detection labels for specific Zernike aberration types and to ensure that different detection labels maintain high response sensitivity to different Zernike aberration terms, after obtaining the K target feature components and their target feature coefficients for each detection label, the aberration sensitivity of each Zernike aberration term is evaluated using the target feature coefficients of each target feature component.

[0099] For example, an aberration sensitivity assessment model is established to quantitatively analyze the extracted target feature coefficients, evaluate the response sensitivity of the target feature coefficients to different Zernike aberration terms, and assign corresponding weight coefficients to each target feature coefficient according to the aberration sensitivity assessment results. Then, the target feature components are fused according to the weight coefficients corresponding to each target feature coefficient under different Zernike aberration terms to construct a (KN)×M-dimensional multi-detection label joint feature vector.

[0100] Among them, the aberration sensitivity assessment can be calculated based on the total target feature coefficient of the target feature component corresponding to each detection mark under different aberration fluctuations, relative to the gradient, correlation or other statistical response indicators of the change of the same Zernike aberration term.

[0101] For example, each detection marker corresponds to 100 target feature components. Based on the target feature coefficients, it is determined that the circular hole has high sensitivity to Zernike aberration terms 5-17, the circular ring has high sensitivity to Zernike aberration terms 13-27, and the square ring has high sensitivity to Zernike aberration terms 21-37. Therefore, the weight coefficients of the target feature components of the circular hole, circular ring, and square ring in the Zernike aberration term 5 are assigned as follows: 0.9, 0.05, 0.05; in the Zernike aberration term 17: 0.5, 0.5, 0; in the Zernike aberration term 27: 0, 0.5, 0.5; in the Zernike aberration term 37: 0.1, 0.1, 0.8, and so on. Next, the target feature components corresponding to each detection label under the same Zernike aberration term are fused according to the weight coefficients to obtain a 300*33 dimensional joint feature vector of multiple detection labels. Of course, after assigning weight coefficients to each detection label in each Zernike aberration term, the weight allocation for each target feature component under that detection label can be further refined based on the assigned weight coefficients. For example, the weight coefficient (0.9) of the circular hole in the 5th Zernike aberration term can be used to further refine the weight allocation for the 100 target feature components corresponding to the circular hole. It should be noted that the above data is only an example; the specific number of target feature components and the assigned weight coefficient values ​​can be set according to actual needs.

[0102] Furthermore, based on the target feature coefficients, the joint feature vector of multiple detection labels, and the Zernike aberration coefficient samples, a mapping relationship model is constructed to utilize the joint feature vector of multiple detection labels to enable each detection label to maintain a high response sensitivity to different Zernike aberration terms, effectively overcoming the limitation of a single detection label being insufficient to cover multiple types of aberrations.

[0103] Since different detection markers have different sensitivities to different aberration terms, the multi-marker joint feature vector formed by fusion can integrate the complementary information of multiple detection markers, significantly enhance the representation ability of multiple Zernike aberration terms, and improve the overall inversion model's ability to identify complex wave aberrations and its stability.

[0104] In one specific embodiment, step S141 includes:

[0105] S1411. Obtain the first feature coefficient of the same detection marker under the first Zernike aberration coefficient sample and the second feature coefficient under the second Zernike aberration coefficient sample;

[0106] The target Zernike aberration coefficients of the first Zernike aberration coefficient sample and the second Zernike aberration coefficient sample are different.

[0107] S1412. Calculate the first variation range of the first feature coefficient and the second feature coefficient, and the second variation range of the target Zernike aberration term coefficient of the first Zernike aberration coefficient sample and the second Zernike aberration coefficient sample;

[0108] S1413. Determine the aberration sensitivity of the detection marker to the Zernike aberration term of the target based on the first and second variation amplitudes.

[0109] Specifically, for the same detection marker, the target feature coefficients corresponding to each target feature component of the detection marker will vary with the Zernike aberration coefficients. By measuring the response amplitude of the target feature coefficients of each target feature component under each set of Zernike aberration coefficient samples, the aberration sensitivity of the detection marker to each Zernike aberration term can be determined.

[0110] In one possible implementation, using two sets of Zernike aberration coefficient samples that differ only in the target Zernike aberration coefficient, the first and second feature coefficients of each target feature component of the detection marker are calculated respectively. Then, the total coefficient change range of the first and second feature coefficients (first change range) and the coefficient change range of the two sets of Zernike aberration coefficient samples in the target Zernike aberration term (second change range) are calculated. If the difference between the first change range and the second change range meets a preset value, it indicates that the detection marker has high aberration sensitivity to the target Zernike aberration term; otherwise, it indicates that the detection marker has low aberration sensitivity to the target Zernike aberration term.

[0111] For example, for the detection marker of a circular hole, if only the fifth Zernike aberration coefficient differs between the two sets of Zernike aberration coefficient samples, while the coefficients of the other Zernike aberration items are the same, and the five Zernike aberration coefficients of the two sets of Zernike aberration coefficient samples differ by 10%, and the total feature coefficients of each target feature component in the two sets of Zernike aberration coefficient samples differ by no less than 50%, then it indicates that the circular hole detection marker has high sensitivity to the aberration of the fifth Zernike aberration item.

[0112] This specific implementation method performs targeted optimization design of detection markers for specific Zernike aberration terms, so that different detection markers maintain high response sensitivity to different Zernike aberration terms, thereby overcoming the limitation that a single detection marker cannot take into account multiple types of aberrations.

[0113] In one specific embodiment, step S13 further includes:

[0114] S131. Perform data dimensionality reduction processing on the simulated defocused image set corresponding to each detection mark to obtain the candidate feature components and their candidate feature coefficients corresponding to each detection mark.

[0115] S132. Determine the third aberration sensitivity of each candidate feature component to each Zernike aberration term based on the candidate feature coefficients.

[0116] S133. Filter candidate feature components whose third aberration sensitivity is greater than a preset threshold to obtain the target feature components and their target feature coefficients.

[0117] Specifically, after performing dimensionality reduction on the simulated defocused image set corresponding to each detection marker, S candidate feature components and their candidate feature coefficients can be extracted, where S is a positive integer and S≥K. However, some feature components are sensitive to Zernike aberration terms, while others are not. Therefore, in order to effectively eliminate redundant information and low-sensitivity features, enhance the robustness of the aberration inversion process, improve computational efficiency, and reduce the impact of noise and disturbances on the measurement results, the third aberration sensitivity of each candidate feature component to each Zernike aberration term can be determined based on the candidate feature coefficients. Candidate feature components with low third aberration sensitivity are eliminated, while candidate feature components with high third aberration sensitivity are retained as target feature components for subsequent mapping relationship model construction.

[0118] For example, using two sets of Zernike aberration coefficient samples, the candidate feature coefficients of each candidate feature component are calculated. The candidate feature coefficients of the same candidate feature component under each Zernike aberration term are compared to determine the sensitivity of each candidate feature component to the third aberration of each Zernike aberration term. If the sensitivity of the candidate feature component to the third aberration of each Zernike aberration term is low, the candidate feature component can be determined as a low-sensitivity feature and can be eliminated. Conversely, if the sensitivity of the candidate feature component to the third aberration of one or more Zernike aberration terms is high, the candidate feature component can be determined as the target feature component.

[0119] This specific implementation effectively eliminates redundant information and low-sensitivity features by performing sensitivity screening on candidate feature components, thereby enhancing the robustness of the aberration inversion process, improving computational efficiency, and reducing the impact of noise and disturbances on the measurement results.

[0120] In one specific implementation, step S14 includes: using a fitting algorithm to fit the target feature coefficients and Zernike aberration coefficient samples to obtain a mapping relationship model.

[0121] Specifically, a fitting algorithm can be used to fit the target feature coefficients to each Zernike aberration coefficient sample. Using the target feature coefficients as input and the Zernike aberration coefficients in each Zernike aberration coefficient sample as ground truth labels, a mapping relationship model between the target feature coefficients and the Zernike aberration coefficients can be obtained. This mapping relationship model can be established using at least one of multiple linear regression, partial least squares regression, or artificial neural networks. The fitting algorithm can be multiple linear regression (MLR), partial least squares regression (PLSR), artificial neural network (ANN), or other regression algorithms that can be used to establish nonlinear or linear mapping relationships.

[0122] In a specific example, Principal Component Analysis (PCA) is used for image feature dimensionality reduction, and a linear fitting method is used to establish a mapping model between the target feature coefficients and Zernike wavelet aberration coefficients. To verify the effectiveness of this embodiment, a comparative analysis is performed between the PCA aberration measurement based on multi-detector joint analysis (MPAM-PCA) and the PCA aberration measurement based on multi-directional one-dimensional markers (AMAI-PCA). Figure 6 As shown in the figure, the input aberration, the PCA aberration measurement results based on multi-directional one-dimensional markers, and the PCA aberration measurement results based on joint analysis of multiple detection markers in this embodiment are presented. The results show that the root mean square error of wavelet aberrations obtained analytically based on multi-directional one-dimensional markers is 3.1 mλ, while the root mean square error in this embodiment is reduced to 0.7 mλ. Figure 7 As shown, the error comparison results of the two methods are further presented. The maximum error based on multi-directional one-dimensional marking is 8.2mλ, while the maximum error of the embodiment is only 1.6mλ.

[0123] The above results show that this embodiment achieves comprehensive and efficient sampling of the pupil by optimizing the detection markers for specific aberration types and combining multiple detection markers with complementary spatial frequency response characteristics. This can significantly improve the inversion accuracy and stability of more Zernike aberration terms, especially higher-order aberration terms. At the same time, this embodiment can perform simultaneous acquisition of multiple marker images under a single illumination mode, which significantly improves the sampling efficiency of marker images and the efficiency of in-situ measurement, and does not require hardware modification of existing equipment.

[0124] This embodiment achieves comprehensive and efficient sampling of the pupil by combining multiple detection markers with complementary spatial frequency response characteristics, significantly improving the measurability and measurement accuracy of more Zernike aberration terms, especially higher-order aberration terms. It can also perform targeted optimization design of detection markers for specific Zernike aberration types, so that different detection markers maintain high response sensitivity to different aberration terms, thereby overcoming the limitation of a single detection marker not being able to cover multiple types of aberrations. At the same time, by performing dimensionality reduction decomposition and sensitivity screening on feature coefficients, redundant information and low-sensitivity features are effectively eliminated, enhancing the robustness of the aberration inversion process, improving computational efficiency, and reducing the impact of noise and disturbances on the measurement results.

[0125] Example 2

[0126] In one specific embodiment of this disclosure, an in-situ measurement method for objective wavelet aberration of an optical quantity detection device is provided, such as... Figure 8 As shown, the in-situ measurement method for objective wavelet aberration includes:

[0127] S21. Multiple detection marks are set up within the same field of view of the optical quantity detection equipment to be measured;

[0128] S22, drive the optical quantity detection device to acquire a set of actual multi-marker defocused images at multiple defocused positions;

[0129] S23. Perform dimensionality reduction processing on the actual multi-marker defocused image set to obtain the feature components of each target and their actual feature coefficients;

[0130] S24. Input the actual characteristic coefficients into the mapping relationship model to obtain the actual Zernike aberration coefficients of the objective lens in the optical quantity detection device to be measured;

[0131] The mapping relationship model is pre-built based on the construction method of Example 1.

[0132] Specifically, in actual measurement, step S21 involves placing a test mask or test wafer containing various detection marks on the workpiece stage of the optical quantity detection equipment, ensuring that multiple detection marks are located within the same field of view. For example, the test wafer may contain various detection marks consistent with the constructed mapping model, including "circular holes," "circular rings," and "square rings." Of course, this can be further extended to other combinations of detection marks with different spatial frequency response characteristics, depending on the type of aberration to be measured and the imaging characteristics of the equipment. This specific embodiment does not limit this approach.

[0133] Step S22 drives the existing area array detector on the optical quantity under test to simultaneously acquire a set of actual multi-marker defocused images of all test marks within the field of view at various defocus positions. During the measurement process, the illumination mode of the optical quantity under test is kept constant, and the objective lens or workpiece stage is controlled to perform defocus scanning along the Z-axis under a single illumination condition. By simultaneously acquiring information on multiple test marks within the same field of view and under a single illumination mode, the image acquisition process is significantly optimized, improving mark image sampling efficiency and overall measurement efficiency, while avoiding additional errors introduced by frequent switching of measurement conditions.

[0134] Step S23 involves performing dimensionality reduction on the acquired set of actual multi-labeled defocused images and projecting them onto the feature decomposition basis established during the construction of the mapping relationship model, in order to extract the target feature components and their actual feature coefficients consistent with the constructed mapping relationship model. The dimensionality reduction process can be found in the aforementioned embodiments and will not be repeated here.

[0135] Finally, in step S24, the actual characteristic coefficients are input into the pre-built mapping relationship model to invert and obtain the Zernike wave aberration coefficients of the objective lens of the current optical quantity detection device, thereby realizing rapid in-situ measurement of objective lens wave aberrations.

[0136] This specific implementation can complete the synchronous image acquisition of multiple detection marks within the same field of view under a single illumination mode, reducing the illumination mode switching and repeated shooting process, optimizing the image measurement process, improving the sampling efficiency of the mark image and the overall measurement efficiency, and eliminating the need for additional hardware modifications to existing optical quantity detection equipment. It can be implemented directly using the existing imaging system and area array detector, has strong compatibility, is easy to integrate into engineering, and can achieve non-invasive, high-precision in-situ measurement of objective wavelet aberrations.

[0137] In one specific embodiment, prior to step S23, the in-situ measurement method of objective wavelet aberration further includes:

[0138] Preprocessing of the actual multi-marker defocused image set;

[0139] Preprocessing includes at least one of normalization, background correction, noise suppression, and image registration.

[0140] Specifically, before performing dimensionality reduction on the actual multi-marker out-of-focus image set, preprocessing such as normalization, background correction, noise suppression, and image registration can be performed on the acquired actual multi-marker out-of-focus image set to improve the signal-to-noise ratio and sharpness of the actual multi-marker out-of-focus image set.

[0141] In one specific embodiment, the optical quantity detection device may include one of the following: a transmission-reflection optical quantity detection device, a total reflection optical quantity detection device, and a reflection optical quantity detection device that includes a vertical incident light path and an oblique incident light path.

[0142] The in-situ measurement method of this specific embodiment does not rely on the modification of the hardware structure of the above-mentioned optical quantity detection equipment. It only requires arranging a variety of pre-designed detection marks on the wafer or mask of the workpiece stage and using the original area array detector of the equipment to collect defocused images, so as to realize the in-situ measurement and inversion of the wavelet aberration of the current objective lens. It has good system compatibility and engineering application convenience.

[0143] In a specific example, such as Figure 9 As shown, in the offline mapping relationship model construction process, multiple sets of Zernike aberration coefficient samples and 1-L sets of measurement modes (multiple detection markers and multiple sets of defocus distances) are input into the vector imaging forward model established according to the actual hardware parameters to obtain a complete set of simulated defocus images corresponding to each set of Zernike aberrations, each detection marker, and each defocus plane. PCA principal component analysis is performed on all simulated defocus image sets to obtain target feature components PC1, PC2...PCL, and corresponding target feature coefficients V1, V2...VL. By evaluating the aberration sensitivity of the target feature coefficients, a joint feature vector of multiple detection markers is constructed. A fitting algorithm is used to fit the target feature coefficients with each Zernike aberration coefficient sample to obtain the mapping relationship model between the target feature coefficients and the Zernike aberration coefficients.

[0144] In actual measurement, the actual multi-marker defocus image set is acquired through the actual objective lens of the optical quantity under test device in the same 1-L group measurement mode. The target feature components PC1, PC2...PCL in the offline stage are directly reused. The actual multi-marker defocus image set is projected onto the target feature components PC1, PC2...PCL to obtain the corresponding actual feature coefficients V1', V2'...VL'. The actual feature coefficients V1', V2'...VL' are input into the mapping relationship model to directly solve the Zernike aberration coefficients of the actual objective lens in reverse.

[0145] This embodiment achieves comprehensive and efficient sampling of the pupil by combining multiple detection markers with complementary spatial frequency response characteristics. This significantly improves the measurability and measurement accuracy of more Zernike aberration terms, especially higher-order aberration terms. Furthermore, it allows for targeted optimization design of detection markers for specific Zernike aberration types, ensuring that different markers maintain high response sensitivity to different aberration terms. This overcomes the limitation of a single detection marker not being able to cover multiple aberration types. Simultaneously, by performing dimensionality reduction decomposition and sensitivity screening on feature coefficients, redundant information and low-sensitivity features are effectively eliminated, enhancing the robustness of the aberration inversion process, improving computational efficiency, and reducing the impact of noise and disturbances on the measurement results. Moreover, simultaneous image acquisition of multiple detection markers within the same field of view can be completed under a single illumination mode, reducing illumination mode switching and repeated shooting processes, optimizing the image measurement process, improving marker image sampling efficiency and overall measurement efficiency. It also eliminates the need for additional hardware modifications to existing optical quantity detection equipment, allowing direct use of existing imaging systems and area array detectors. This provides strong compatibility, facilitates engineering integration, and enables non-invasive, high-precision in-situ measurement of objective wavelet aberrations.

[0146] Example 3

[0147] In one specific embodiment of this disclosure, a system for constructing a mapping relationship model for in-situ measurement of objective wavelet aberrations is provided, such as... Figure 10 As shown, the construction system includes:

[0148] Module 101 is established to build a forward vector imaging model based on the hardware parameters of the optical quantity detection device to be measured.

[0149] The calculation module 102 is used to generate multiple sets of Zernike aberration coefficient samples within a preset aberration fluctuation range, and to use a vector imaging forward model to calculate the simulated defocus image set of various detection markers at each of the Zernike aberration coefficient samples and multiple defocus positions; wherein, the spatial frequency response features of the various detection markers are complementary.

[0150] Processing module 103 is used to perform data dimensionality reduction processing on the simulated defocus image set corresponding to each detection mark, and to obtain several target feature components and their target feature coefficients corresponding to each detection mark.

[0151] The construction module 104 is used to construct a mapping relationship model based on the target feature coefficients and Zernike aberration coefficient samples.

[0152] In one specific embodiment, the construction module 104 includes an evaluation unit and a construction unit;

[0153] The evaluation unit is used to evaluate the aberration sensitivity of each detection label to each Zernike aberration term based on the target feature coefficients, and to fuse the target feature components of each detection label based on the aberration sensitivity evaluation results to construct a multi-detection label joint feature vector corresponding to each Zernike aberration term.

[0154] The construction unit is used to construct the mapping relationship model based on the target feature coefficients, the joint feature vector of the multi-detection labels, and the Zernike aberration coefficient samples.

[0155] In one specific embodiment, the evaluation unit is further configured to acquire a first feature coefficient of the same detection marker under a first Zernike aberration coefficient sample and a second feature coefficient under a second Zernike aberration coefficient sample; wherein the target Zernike aberration coefficients of the first Zernike aberration coefficient sample and the second Zernike aberration coefficient sample are different; calculate a first variation range of the first feature coefficient and the second feature coefficient, and a second variation range of the target Zernike aberration coefficients of the first Zernike aberration coefficient sample and the second Zernike aberration coefficient sample; and determine the aberration sensitivity of the detection marker to the target Zernike aberration term based on the first variation range and the second variation range.

[0156] In one specific implementation, the processing module 103 is specifically used to perform data dimensionality reduction processing on the simulated defocus image set corresponding to each detection mark, to obtain the candidate feature components and their candidate feature coefficients corresponding to each detection mark; to determine the third aberration sensitivity of each candidate feature component to each Zernike aberration term based on the candidate feature coefficients; and to filter the candidate feature components whose third aberration sensitivity is greater than a preset threshold to obtain the target feature component and its target feature coefficients.

[0157] In one specific implementation, the construction module 104 is specifically used to fit the target feature coefficients and Zernike aberration coefficient samples using a fitting algorithm to obtain a mapping relationship model.

[0158] In one specific implementation, the spatial frequency range of multiple detection markers under a single illumination mode fully covers the effective transmission frequency range of the pupil.

[0159] In one specific embodiment, the detection mark includes at least two of the following: a circular hole mark, a circular ring mark, and a square ring mark.

[0160] In one specific implementation, the processing module 103 is specifically used to perform feature extraction and dimensionality reduction decomposition on the simulated defocus image set corresponding to each detection marker using principal component analysis and / or singular value decomposition.

[0161] In one specific implementation, the mapping relationship model is established using at least one of multiple linear regression, partial least squares regression, and artificial neural networks.

[0162] For the system embodiments, since they basically correspond to the method embodiments, the relevant parts can be referred to in the description of the method embodiments. The system embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this disclosure according to actual needs.

[0163] This embodiment achieves comprehensive and efficient sampling of the pupil by combining multiple detection markers with complementary spatial frequency response characteristics. This significantly improves the measurability and measurement accuracy of more Zernike aberration terms, especially higher-order aberration terms. Furthermore, it allows for targeted optimization design of detection markers for specific Zernike aberration types, ensuring that different markers maintain high response sensitivity to different aberration terms. This overcomes the limitation of a single detection marker not being able to cover multiple aberration types. Simultaneously, by performing dimensionality reduction decomposition and sensitivity screening on feature coefficients, redundant information and low-sensitivity features are effectively eliminated, enhancing the robustness of the aberration inversion process, improving computational efficiency, and reducing the impact of noise and disturbances on the measurement results. Moreover, simultaneous image acquisition of multiple detection markers within the same field of view can be completed under a single illumination mode, reducing illumination mode switching and repeated shooting processes, optimizing the image measurement process, improving marker image sampling efficiency and overall measurement efficiency. It also eliminates the need for additional hardware modifications to existing optical quantity detection equipment, allowing direct use of existing imaging systems and area array detectors. This provides strong compatibility, facilitates engineering integration, and enables non-invasive, high-precision in-situ measurement of objective wavelet aberrations.

[0164] Example 4

[0165] In one specific embodiment of this disclosure, an in-situ measurement system for objective wavelet aberration of an optical quantity detection device is provided, such as... Figure 11 As shown, the system includes:

[0166] The acquisition module 201 is used to deploy multiple detection marks within the same field of view of the optical quantity detection device under test, and drive the optical quantity detection device under test to acquire a set of actual multi-marker defocus images at multiple defocus positions.

[0167] The acquisition module 202 is used to perform data dimensionality reduction processing on the actual multi-marker defocus image set to obtain the feature components of each target and their actual feature coefficients;

[0168] Measurement module 203 is used to input the actual characteristic coefficients into the mapping relationship model to obtain the actual Zernike aberration coefficients of the objective lens in the optical quantity detection device under test;

[0169] The mapping relationship model is pre-built based on the construction system in Example 3.

[0170] In one specific embodiment, the system further includes:

[0171] The preprocessing module is used to preprocess the actual multi-marker defocus image set;

[0172] Preprocessing includes at least one of normalization, background correction, noise suppression, and image registration.

[0173] In one specific implementation, the actual multi-marker defocused image set is obtained under a single illumination mode.

[0174] In one specific embodiment, the optical quantity detection device includes one of the following: a transmission-reflection optical quantity detection device, a total reflection optical quantity detection device, and a reflection optical quantity detection device that includes a vertical incident light path and an oblique incident light path.

[0175] For the system embodiments, since they basically correspond to the method embodiments, the relevant parts can be referred to in the description of the method embodiments. The system embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this disclosure according to actual needs.

[0176] This embodiment achieves comprehensive and efficient sampling of the pupil by combining multiple detection markers with complementary spatial frequency response characteristics. This significantly improves the measurability and measurement accuracy of more Zernike aberration terms, especially higher-order aberration terms. Furthermore, it allows for targeted optimization design of detection markers for specific Zernike aberration types, ensuring that different markers maintain high response sensitivity to different aberration terms. This overcomes the limitation of a single detection marker not being able to cover multiple aberration types. Simultaneously, by performing dimensionality reduction decomposition and sensitivity screening on feature coefficients, redundant information and low-sensitivity features are effectively eliminated, enhancing the robustness of the aberration inversion process, improving computational efficiency, and reducing the impact of noise and disturbances on the measurement results. Moreover, simultaneous image acquisition of multiple detection markers within the same field of view can be completed under a single illumination mode, reducing illumination mode switching and repeated shooting processes, optimizing the image measurement process, improving marker image sampling efficiency and overall measurement efficiency. It also eliminates the need for additional hardware modifications to existing optical quantity detection equipment, allowing direct use of existing imaging systems and area array detectors. This provides strong compatibility, facilitates engineering integration, and enables non-invasive, high-precision in-situ measurement of objective wavelet aberrations.

[0177] Example 5

[0178] Figure 12 This is a schematic diagram of the structure of an electronic device according to an example embodiment of the present disclosure. The electronic device includes a memory, a processor, and a computer program stored in the memory and used to run on the processor. When the processor executes the computer program, it implements the method for constructing a mapping relationship model for in-situ measurement of objective wavelet aberration as described in any of the above embodiments, or the method for in-situ measurement of objective wavelet aberration of an optical quantity detection device. Figure 12 The electronic device 50 shown is merely an example and should not impose any limitation on the functionality and scope of use of the embodiments disclosed herein.

[0179] like Figure 12 As shown, the electronic device 50 can be manifested in the form of a general-purpose computing device, such as a server device. The components of the electronic device 50 may include, but are not limited to: at least one processor 51, at least one memory 52, and a bus 53 connecting different system components (including memory 52 and processor 51).

[0180] Bus 53 includes a data bus, an address bus, and a control bus.

[0181] The memory 52 may include volatile memory, such as random access memory (RAM) 521 and / or cache memory 522, and may further include read-only memory (ROM) 523.

[0182] The memory 52 may also include a program tool 525 (or utility) having a set (at least one) program module 524, such program module 524 including but not limited to: an operating system, one or more application programs, other program modules, and program data, each or some combination of these examples may include an implementation of a network environment.

[0183] The processor 51 executes various functional applications and data processing by running computer programs stored in the memory 52, such as the method for constructing a mapping relationship model for in-situ measurement of objective wavelet aberration provided in any of the above embodiments, or the method for in-situ measurement of objective wavelet aberration of an optical quantity detection device.

[0184] Electronic device 50 can also communicate with one or more external devices 54 (e.g., keyboard, pointing device, etc.). This communication can be performed via input / output (I / O) interface 55. Furthermore, electronic device 50 can also communicate with one or more networks (e.g., local area network (LAN), wide area network (WAN), and / or public networks, such as the Internet) via network adapter 56. Figure 12 As shown, network adapter 56 communicates with other modules of electronic device 50 via bus 53. It should be understood that, although not shown in the figure, other hardware and / or software modules can be used in conjunction with electronic device 50, including but not limited to: microcode, device drivers, redundant processors, external disk drive arrays, RAID (disk array) systems, tape drives, and data backup storage systems.

[0185] It should be noted that although several units / modules or sub-units / modules of the electronic device have been mentioned in the detailed description above, this division is merely exemplary and not mandatory. In fact, according to embodiments of this disclosure, the features and functions of two or more units / modules described above can be embodied in one unit / module. Conversely, the features and functions of one unit / module described above can be further divided and embodied by multiple units / modules.

[0186] Example 6

[0187] This disclosure also provides a computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements the method for constructing a mapping relationship model for in-situ measurement of objective wavelet aberrations provided in any of the above embodiments, or the method for in-situ measurement of objective wavelet aberrations of an optical quantity detection device.

[0188] The readable storage medium may be more specifically adopted, including but not limited to: portable disk, hard disk, random access memory, read-only memory, erasable programmable read-only memory, optical storage device, magnetic storage device, or any suitable combination thereof.

[0189] Example 7

[0190] This disclosure also provides a computer program product, including a computer program that, when executed by a processor, implements the method for constructing a mapping relationship model for in-situ measurement of objective wavelet aberration as described above, or the method for in-situ measurement of objective wavelet aberration of an optical quantity detection device.

[0191] The program code for executing the computer program product of this disclosure can be written in any combination of one or more programming languages, and the program code can be executed entirely on a user device, partially on a user device, as a stand-alone software package, partially on a user device and partially on a remote device, or entirely on a remote device.

[0192] While specific embodiments of this disclosure have been described above, those skilled in the art should understand that these are merely illustrative examples, and the scope of protection of this disclosure is defined by the appended claims. Those skilled in the art can make various changes or modifications to these embodiments without departing from the principles and essence of this disclosure, but all such changes and modifications fall within the scope of protection of this disclosure.

Claims

1. A method for constructing a mapping relationship model for in-situ measurement of objective wavelet aberrations, characterized in that, The construction method includes: A vector imaging forward model is established based on the hardware parameters of the optical quantity detection device to be measured. Multiple sets of Zernike aberration coefficient samples are generated within a preset aberration fluctuation range. The vector imaging forward model is used to calculate the simulated defocus image set of various detection markers at each of the Zernike aberration coefficient samples and multiple defocus positions. The spatial frequency response features of the various detection markers are complementary. Data dimensionality reduction processing is performed on the simulated defocused image set corresponding to each detection marker to obtain several target feature components and their target feature coefficients corresponding to each detection marker. A mapping relationship model is constructed based on the target feature coefficients and the Zernike aberration coefficient samples.

2. The construction method according to claim 1, characterized in that, The step of constructing a mapping relationship model based on the target feature coefficients and the Zernike aberration coefficient samples includes: Based on the target feature coefficients, the aberration sensitivity of each detection label to each Zernike aberration term is evaluated, and the target feature components of each detection label are fused based on the aberration sensitivity evaluation results to construct a multi-detection label joint feature vector corresponding to each Zernike aberration term. The mapping relationship model is constructed based on the target feature coefficients, the joint feature vector of the multi-detection labels, and the Zernike aberration coefficient samples.

3. The construction method according to claim 2, characterized in that, The step of evaluating the aberration sensitivity of each detection marker to each Zernike aberration term based on the target feature coefficients includes: Obtain the first feature coefficient of the same detection marker under the first Zernike aberration coefficient sample and the second feature coefficient under the second Zernike aberration coefficient sample; The target Zernike aberration coefficients of the first Zernike aberration coefficient sample and the second Zernike aberration coefficient sample are different. Calculate the first variation magnitude of the first feature coefficient and the second feature coefficient, and the second variation magnitude of the target Zernike aberration term coefficient of the first Zernike aberration coefficient sample and the second Zernike aberration coefficient sample; The aberration sensitivity of the detection marker to the target Zernike aberration term is determined based on the first change magnitude and the second change magnitude.

4. The construction method according to any one of claims 1 to 3, characterized in that, The step of performing data dimensionality reduction processing on the simulated defocus image set corresponding to each detection marker to obtain several target feature components and their target feature coefficients corresponding to each detection marker includes: Perform dimensionality reduction processing on the simulated defocused image set corresponding to each detection marker to obtain the candidate feature components and their candidate feature coefficients corresponding to each detection marker. The third aberration sensitivity of each candidate feature component to each Zernike aberration term is determined based on the candidate feature coefficients. The candidate feature components whose third aberration sensitivity is greater than a preset threshold are selected to obtain the target feature components and their target feature coefficients; And / or, The step of constructing a mapping relationship model based on the target feature coefficients and the Zernike aberration coefficient samples includes: The target feature coefficients and the Zernike aberration coefficient samples are fitted using a fitting algorithm to obtain the mapping relationship model; And / or, The spatial frequency range of the various detection markers under a single illumination mode fully covers the effective transmission frequency range of the pupil; And / or, The detection markings include at least two of the following: circular hole markings, circular ring markings, and square ring markings; And / or, The step of performing data dimensionality reduction processing on the simulated off-focus image set corresponding to each of the detection markers includes: Principal component analysis and / or singular value decomposition are used to extract features and reduce dimensions of the simulated defocused image set corresponding to each detection marker. And / or, The mapping relationship model is established using at least one of multiple linear regression, partial least squares regression, and artificial neural networks.

5. A method for in-situ measurement of objective wavelet aberration in an optical quantity detection device, characterized in that, The method includes: Multiple detection marks are set up within the same field of view of the optical quantity detection equipment to be measured; The device for detecting the optical quantity under test is driven to acquire a set of actual multi-marker defocused images at multiple defocused positions; The actual multi-marker defocused image set is subjected to data dimensionality reduction processing to obtain the feature components of each target and their actual feature coefficients; The actual characteristic coefficients are input into the mapping relationship model to obtain the actual Zernike aberration coefficients of the objective lens in the optical quantity detection device to be measured; The mapping relationship model is pre-constructed based on the construction method of any one of claims 1 to 4.

6. The method for in-situ measurement of objective wavelet aberration according to claim 5, characterized in that, Before the step of performing data dimensionality reduction processing on the actual multi-marker off-focus image set, the method further includes: The actual multi-marker defocus image set is preprocessed; The preprocessing includes at least one of normalization, background correction, noise suppression, and image registration; And / or, the actual multi-marker defocus image set is obtained in a single illumination mode; And / or, The optical quantity detection equipment includes one of the following: a transmission-reflection optical quantity detection equipment, a total reflection optical quantity detection equipment, and a reflection optical quantity detection equipment that includes a vertical incident light path and an oblique incident light path.

7. A system for constructing a mapping relationship model for in-situ measurement of objective wavelet aberrations, characterized in that, The construction system includes: A module is established to build a forward vector imaging model based on the hardware parameters of the optical quantity to be measured detection device. The calculation module is used to generate multiple sets of Zernike aberration coefficient samples within a preset aberration fluctuation range, and to use the vector imaging forward model to calculate the simulated defocus image sets of various detection markers at each of the Zernike aberration coefficient samples and multiple defocus positions; wherein, the spatial frequency response features of the various detection markers are complementary. The processing module is used to perform data dimensionality reduction processing on the simulated defocus image set corresponding to each detection mark, and obtain several target feature components and their target feature coefficients corresponding to each detection mark; The module is used to construct a mapping relationship model based on the target feature coefficients and the Zernike aberration coefficient samples.

8. An in-situ measurement system for objective wavelet aberration of an optical quantity detection device, characterized in that, The system includes: The acquisition module is used to deploy multiple detection marks within the same field of view of the optical quantity detection device under test, and drive the optical quantity detection device under test to acquire a set of actual multi-marker defocus images at multiple defocus positions; The acquisition module is used to perform data dimensionality reduction processing on the actual multi-marker defocus image set to obtain each target feature component and its actual feature coefficients; The measurement module is used to input the actual characteristic coefficients into the mapping relationship model to obtain the actual Zernike aberration coefficients of the objective lens in the optical quantity detection device to be measured. The mapping relationship model is pre-built based on the construction system described in claim 7.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and for running on the processor, characterized in that, When the processor executes the computer program, it implements the method for constructing the mapping relationship model for in-situ measurement of objective wavelet aberration as described in any one of claims 1 to 4, or the method for in-situ measurement of objective wavelet aberration of the optical quantity detection device as described in claim 5 or 6.

10. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by the processor, it implements the method for constructing the mapping relationship model for in-situ measurement of objective wavelet aberration as described in any one of claims 1 to 4, or the method for in-situ measurement of objective wavelet aberration of the optical quantity detection device as described in claim 5 or 6.

Images