Visible light near-infrared metasurface unit partition exposure control method

Abstract

The present disclosure provides a visible light near-infrared super surface unit partition exposure control method, comprising: binding the target complex response of the super surface unit to the unit position according to the working wavelength state and the polarization state to obtain a response constraint set; obtaining a geometric target graph and a geometric tolerance graph by inversion from the response constraint set; inputting the geometric target graph and the geometric tolerance graph into an exposure transfer model to obtain a local sensitivity graph and a neighborhood coupling matrix; obtaining a partition index graph and a boundary coupling set by partitioning from the local sensitivity graph and the neighborhood coupling matrix; obtaining a partition exposure parameter, a predicted geometric contour, a two-domain residual error, a converged exposure parameter and an exposure execution sequence by solving from the partition index graph, the boundary coupling set, the geometric target graph and the geometric tolerance graph; obtaining an exposure execution deviation from the difference between the execution value of the exposure device and the exposure execution sequence; correcting the exposure execution sequence of the unexposed partition by the exposure execution deviation to obtain a visible light near-infrared super surface pattern.

Description

Technical Field

[0001] This invention relates to the field of micro-nano optical manufacturing, and in particular to a method for controlling the zoned exposure of visible and near-infrared metasurface units. Background Technology

[0002] Visible and near-infrared dual-band imaging plays a dual role in daytime imaging and night vision illumination in fields such as machine vision, biomedical detection, facial recognition, and vehicle sensing. Traditional implementations rely on cascaded multiple refractive lenses, along with filters and polarization elements, to achieve band separation and polarization control, resulting in high overall thickness and assembly complexity. Metasurfaces, composed of subwavelength-scale dielectric pillar arrays, allow each unit to modulate the amplitude and phase of transmitted light based on its own geometric scale and contour direction, performing focusing, achromatic aberration, and polarization separation on a single plane. By extending the operating band of metasurfaces from visible light to near-infrared, a single device can simultaneously provide the designed complex response in both the visible imaging channel near 532 nm and the infrared illumination channel near 850 nm, significantly reducing the thickness and weight of the optical system. The core of such devices lies in the control of the complex transmission coefficient unit by unit. The unit position, critical size, and contour direction jointly determine the cross-band and cross-polarization complex response distribution, thus requiring significantly higher precision in pattern transfer during manufacturing than conventional micro / nano structures.

[0003] Mass production of visible-near-infrared metasurfaces relies on photolithography, a process that includes optical response design, unit geometry inversion, exposure pattern preparation, and etching transfer. Currently, the design and manufacturing processes are largely separate. The design phase uses angular spectrum propagation and rigorous coupled-wave analysis to determine unit geometry parameters, with the output geometry based on an ideal profile. The manufacturing phase employs the same uniform exposure conditions used in integrated circuits, meaning the entire substrate uses a single exposure dose and a single scanning trajectory, with dose compensation relying on the optical proximity effect at the mask end. However, the morphological differences within a metasurface array on the same wafer are far greater than in conventional circuit patterns: the focusing phase distribution causes a continuous change in unit geometry from the center to the edge, altering the arrangement density, and resulting in significant differences in neighborhood coupling strength between dense and sparse regions. Under uniform exposure conditions, the energy accumulation from the same dose differs between dense and sparse regions, causing pattern transfer accuracy to drift along the wafer's position. Existing correction methods rely solely on geometric size deviations, failing to reflect optical response deviations back to the exposure parameters, and leaving cross-band and cross-polarization complex response errors unconstrained. The drift between the actual dose and the actual scanning position relative to the command value at the device end also lacks a correction channel to be transmitted to the incomplete area.

[0004] The technical defects exposed by the above situation are as follows: the non-uniformity of exposure energy transfer cannot be expressed and quantitatively compensated in different areas; it is difficult to form consistent pattern transfer accuracy under unified exposure conditions; the critical size offset of the unit and the distortion of the boundary contour accumulate along the position within the film; geometric deviation and optical deviation are in separate decision links; cross-band complex response deviation lacks a reverse correction path; and the pattern consistency and complex response accuracy of visible and near-infrared metasurfaces are limited. Summary of the Invention

[0005] Based on this, it is necessary to propose a visible light near-infrared metasurface unit zoned exposure control method to address the aforementioned technical problems.

[0006] The present invention adopts the following technical solution.

[0007] The first aspect of this invention discloses a method for controlling the zonal exposure of visible-near-infrared metasurface units, comprising: binding the target complex response of the metasurface unit to the unit position according to the working wavelength state and polarization state to obtain a response constraint set; inverting the response constraint set to obtain a geometric target map and a geometric tolerance map; inputting the geometric target map and the geometric tolerance map into an exposure transfer model to obtain a local sensitivity map and a neighborhood coupling matrix; dividing the local sensitivity map and the neighborhood coupling matrix to obtain a zonal index map and a boundary coupling set; solving for the zonal exposure parameters from the zonal index map, the boundary coupling set, the geometric target map, and the geometric tolerance map; inputting the zonal exposure parameters into the exposure transfer model to obtain a predicted geometric profile; obtaining a two-domain residual from the predicted geometric profile, the geometric target map, and the response constraint set; correcting the zonal exposure parameters with the two-domain residual to obtain converged exposure parameters; obtaining an exposure execution sequence from the converged exposure parameters, the zonal index map, and the boundary coupling set; obtaining an exposure execution deviation from the difference between the execution value of the exposure device and the exposure execution sequence; and correcting the exposure execution sequence of the unexposed zonals with the exposure execution deviation to obtain a visible-near-infrared metasurface pattern.

[0008] Furthermore, the geometric target map and geometric tolerance map are obtained by inverting the response constraint set. Specifically, this includes: discretizing the geometric vectors of the metasurface units according to the manufacturing resolution limit; solving for the complex transmission coefficient in the structure defined by the material level parameters to obtain a forward mapping from the geometric vector to the complex transmission coefficient; forming an inversion objective function by the difference between the target complex response in the response constraint set and the forward mapping output; correcting the geometric vector within the value defined by the manufacturing resolution limit and selecting the geometric vector that decreases in the inversion objective function; writing the unit number, the geometric vector, and the contour template into the geometric target map; perturbing the geometric vector according to the size adjustment step size, obtaining the change in the complex transmission coefficient from the forward mapping, and forming a sensitivity matrix; dividing the preset complex response error threshold by the sensitivity norm of the individual geometric parameters in the sensitivity matrix to obtain the parameter allowable deviation; and writing the unit number and the parameter allowable deviation into the geometric tolerance map.

[0009] Furthermore, the target complex response of the metasurface unit is bound to the unit position according to the working wavelength state and polarization state to obtain the response constraint set. Specifically, this includes: transforming the target light field of the set focal plane to the spatial spectrum; applying the inverse propagation transfer function to the spatial spectrum and inversely transforming it to obtain the complex light field of the metasurface plane under the working wavelength state; substituting the polarization basis vector into the model describing the vector transformation between the incident electric field and the output electric field, and extracting the complex transmission coefficient under the polarization state from the complex light field of the metasurface plane; performing maximum modulus normalization on the complex transmission coefficient to obtain the target amplitude; converting the argument of the complex transmission coefficient to the set phase period to obtain the target phase; the target amplitude and the target phase constitute the target complex response; calculating the unit number according to the center coordinates and period parameters of the metasurface array; forming a state index from the working wavelength state and polarization state; forming a constraint record from the unit number, state index, and target complex response; and the set of constraint records constitutes the response constraint set.

[0010] Furthermore, the geometric target map and geometric tolerance map are input into the exposure transfer model to obtain a local sensitivity map and a neighborhood coupling matrix. Specifically, the exposure transfer model configures the exposure pattern according to the unit position and contour template in the geometric target map; the exposure transfer model performs projection energy diffusion calculation on the exposure dose of the metasurface unit to obtain the substrate energy field; performs photosensitive material state transformation on the substrate energy field to obtain the latent image contour; performs pattern transfer deviation mapping on the latent image contour to obtain the predicted geometric contour; the metasurface unit to which the dose perturbation is applied is denoted as the perturbed unit; the metasurface unit adjacent to the perturbed unit is denoted as the adjacent unit; positive and negative perturbations are applied to the exposure dose of the perturbed unit; the local sensitivity map is formed by the change intensity of the critical size of the perturbed unit relative to the positive and negative perturbations; and the neighborhood coupling matrix is ​​formed by the change intensity of the critical size of the adjacent unit relative to the positive and negative perturbations.

[0011] Furthermore, a partition index map and a boundary coupling set are obtained by dividing the local sensitivity map and the neighborhood coupling matrix. Specifically, this includes: using metasurface units as nodes, and using the dimensionless value synthesized from the bidirectional transmission intensity in the neighborhood coupling matrix, the sensitivity difference in the local sensitivity map, and the node center distance as connection weights to construct a weighted adjacency graph; the sensitivity dispersion is the ratio of the absolute value of the standard deviation of sensitivity within the node set to the absolute value of the mean sensitivity; recursively dividing the node set of the weighted adjacency graph according to the sensitivity dispersion threshold within the partition to obtain candidate partitions; the endpoints of an edge whose connection weight is higher than the inter-partition coupling threshold and belongs to different candidate partitions are marked as the first boundary node and the second boundary node; comparing the normalized cut cost of moving the first boundary node into the candidate partition to which the second boundary node belongs with the normalized cut cost of moving the second boundary node into the candidate partition to which the first boundary node belongs, and performing the move with the larger cost reduction to obtain the exposure control partition; writing the exposure control partition number to which the metasurface unit belongs into the partition index map; and writing the first boundary node number, the second boundary node number, and the connection weight of the edge into the boundary coupling set.

[0012] Furthermore, the zonal exposure parameters are obtained by solving the zonal index map, boundary coupling set, geometric target map, and geometric tolerance map. Specifically, the zonal exposure parameters include: reference dose, boundary compensation amount, and scan connection position; the exposure control zonal is taken from the zonal index map; the reference dose is the main dose within the exposure control zonal; the exposure control zonal connected by a record in the boundary coupling set is denoted as the first zonal and the second zonal; the boundary compensation amount is the dose correction value applied to the boundary band between the first zonal and the second zonal; the scan connection position is the boundary coordinate between the scan trajectory of the first zonal and the scan trajectory of the second zonal. The exposure transfer model converts the reference dose, boundary compensation amount, and scan connection position into iterative geometric parameters. The objective function is solved by the difference between the iterative geometric parameters and the target geometric parameters in the geometric target map, normalized by the geometric tolerance map; the ratio of the boundary compensation amount to the sum of the reference doses of the first and second zones, weighted by the connection weight of the record; and the offset of the scan connection position relative to the midpoint of the boundary. The reference dose, boundary compensation amount, and scan connection position are corrected, and the reference dose, boundary compensation amount, and scan connection position that are within the range defined by the geometric tolerance map are taken as the zone exposure parameters.

[0013] Furthermore, the dual-domain residual is obtained from the predicted geometric profile, the geometric target map, and the response constraint set. Specifically, this includes: extracting the geometric parameters of the metasurface unit from the predicted geometric profile; the difference between these geometric parameters and the target geometric parameters in the geometric target map forms the geometric difference; normalizing the geometric difference using a geometric tolerance map to obtain the geometric residual; reconstructing the complex transmission coefficient of the metasurface unit in the working wavelength state and polarization state from the predicted geometric profile, denoted as the reconstructed complex response; the difference between the reconstructed complex response and the target complex response in the response constraint set forms the optical residual; and combining the geometric residual and the optical residual according to a preset weight to obtain the dual-domain residual.

[0014] The second aspect of this invention discloses a visible-near-infrared metasurface unit partitioned exposure control system, applied to the method described in the first aspect, comprising: a response constraint module, which binds the target complex response of the metasurface unit to the unit position according to the working wavelength state and polarization state to obtain a response constraint set; a geometric inversion module, which inverts the response constraint set to obtain a geometric target map and a geometric tolerance map; an exposure transfer module, which inputs the geometric target map and the geometric tolerance map into an exposure transfer model to obtain a local sensitivity map and a neighborhood coupling matrix; a partitioning module, which partitions the local sensitivity map and the neighborhood coupling matrix to obtain a partition index map and a boundary coupling set; and a parameter solving module. The partitioned exposure parameters are obtained by solving the partitioned index map, boundary coupling set, geometric target map, and geometric tolerance map. The residual correction module inputs the partitioned exposure parameters into the exposure transfer model to obtain the predicted geometric profile. The predicted geometric profile, geometric target map, and response constraint set are used to obtain the two-domain residual. The two-domain residual is used to correct the partitioned exposure parameters to obtain the converged exposure parameters. The exposure execution module obtains the exposure execution sequence from the converged exposure parameters, partitioned index map, and boundary coupling set. The difference between the execution value of the exposure device and the exposure execution sequence is used to obtain the exposure execution deviation. The exposure execution deviation is used to correct the exposure execution sequence of the unexposed partition to obtain the visible light near-infrared metasurface pattern.

[0015] A third aspect of the present invention discloses a terminal, including a processor and a storage medium; characterized in that:

[0016] The storage medium is used to store instructions;

[0017] The processor is configured to operate according to the instructions to perform the steps of the method described in the first aspect.

[0018] A fourth aspect of the present invention discloses a computer-readable storage medium having a computer program stored thereon, characterized in that the program, when executed by a processor, implements the steps of the method described in the first aspect.

[0019] The beneficial effects of the present invention are as follows: Compared with the prior art, the present invention has the following advantages:

[0020] (1) This invention uses local sensitivity maps and neighborhood coupling matrices to characterize the differences in geometric scale, arrangement density, and neighborhood coupling strength of metasurface units, and divides the exposure control zones accordingly. The uniform exposure conditions are replaced by reference dose, boundary compensation amount, and scan connection position configured according to the zones. The response dispersion within a zone is constrained by the sensitivity dispersion threshold, and the energy diffusion across zones is constrained by the boundary coupling set and boundary compensation amount. The unit critical size offset and boundary contour distortion caused by the non-uniformity of exposure energy transfer are suppressed, and the pattern transfer accuracy tends to be consistent across the entire metasurface.

[0021] (2) This invention combines geometric residuals and optical residuals into dual-domain residuals, and corrects the zoned exposure parameters in reverse along the local sensitivity map and the neighborhood coupling matrix, incorporating manufacturing deviations and cross-band complex response deviations into the same correction loop. The geometric tolerance map provides a normalized scale for the geometric residuals, and the response constraint set provides a cross-band and cross-polarization comparison benchmark for the optical residuals. The exposure execution deviation further corrects the energy command and scan command of the unexposed zone, forming a closed-loop control relationship between optical response, unit geometry, zoned energy, and exposure execution, thus converging the cross-band complex response deviation. Attached Figure Description

[0022] To more clearly illustrate the technical solutions in this invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0023] Figure 1 This is a schematic diagram of the exposure transfer model.

[0024] Figure 2 This is a schematic diagram of the exposure control zone division framework.

[0025] Figure 3 This is a schematic diagram of a visible light near-infrared metasurface unit zoned exposure control system. Detailed Implementation

[0026] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0027] This invention discloses a visible-near-infrared metasurface unit partitioned exposure control method. Using the cross-band complex response of a transmissive metasurface as the control target, steps 1 to 7 constitute a data chain. Step 1 forms a response constraint set. Step 2 inverts the response constraint set to obtain a geometric target map and a geometric tolerance map. Step 3 inputs the geometric target map and geometric tolerance map into an exposure transfer model and outputs a local sensitivity map and a neighborhood coupling matrix. Step 4 partitions the local sensitivity map and neighborhood coupling matrix to obtain a partitioned index map and a boundary coupling set. Step 5 solves for the partitioned exposure parameters using the partitioned index map, boundary coupling set, geometric target map, and geometric tolerance map. Step 6 inputs the partitioned exposure parameters into the exposure transfer model to obtain a predicted geometric profile and obtains converged exposure parameters through dual-domain residual correction. Step 7 obtains the exposure execution sequence from the converged exposure parameters, partitioned index map, and boundary coupling set, and corrects the exposure execution sequence of the unexposed partitions using exposure execution deviation, thus obtaining the visible-near-infrared metasurface pattern.

[0028] The following embodiments use a transmissive metasurface with operating wavelengths of 532 nm and 850 nm, incident polarization of x and y, and a target focused by a predetermined focal plane. A metasurface unit refers to a rectangular dielectric pillar occupying one periodic position in the metasurface array. The operating wavelength state refers to a sampling wavelength within the operating band of the metasurface, and the polarization state refers to a vector direction of the incident electric field. The difference between the two is that the former defines the cross-band sampling position, while the latter defines the vector response direction of the unit.

[0029] Step 1: Bind the target complex response of the metasurface element to the element location according to the operating wavelength state and polarization state to obtain the response constraint set. The target complex response refers to the transmission complex amplitude that a metasurface element needs to provide in one operating wavelength state and one polarization state, which consists of the target amplitude and the target phase. The response constraint set refers to the set of all constraint records.

[0030] Sub-step 1.1: Transform the target light field of the set focal plane to the spatial spectrum, apply the inverse propagation transfer function to the spatial spectrum, and inversely transform to obtain the complex light field of the metasurface plane under the working wavelength state.

[0031] Specifically, this process employs an angular spectrum propagation model. The target light field at the focal plane is subjected to a two-dimensional Fourier transform to obtain its spatial spectrum. This spatial spectrum is then multiplied by the inverse propagation transfer function and subjected to a two-dimensional inverse Fourier transform to output the complex light field of the metasurface plane at two working wavelengths: 532 nm and 850 nm. The complex light field of the metasurface plane provides the required amplitude and phase at each spatial location on the plane containing the metasurface. It differs from the target light field at the focal plane in that it is located on the plane containing the metasurface.

[0032] Sub-step 1.2: Substitute the polarization basis vector into the model describing the vector transformation between the incident electric field and the output electric field, extract the complex transmission coefficient under polarization state from the complex light field of the metasurface plane, normalize it to obtain the target amplitude, and calculate the argument to obtain the target phase.

[0033] Specifically, the vector transformation model uses the Jones matrix to describe the linear transformation from the incident electric field to the emitted electric field. Substituting the x-polarization and y-polarization basis vectors into this model, the complex transmission coefficients in the x-polarization and y-polarization states are output. The complex transmission coefficient is the ratio of the emitted complex amplitude to the incident complex amplitude. Maximum modulus normalization is performed on the complex transmission coefficients within a state, and the target amplitude is output. The argument of the complex transmission coefficients is converted to a set phase period, and the target phase is output. The target complex response is written as: ; in, This represents the target's complex transmission coefficient. Indicates the target amplitude. Indicates the target phase.

[0034] Sub-step 1.3: Calculate the cell number according to the center coordinates and periodic parameters of the metasurface array, form a state index by combining the working wavelength state and polarization state, and combine the cell number, state index and target complex response to form a constraint record.

[0035] Specifically, the element number refers to the unique identifier of a rectangular dielectric column in the array row and column position, calculated from the array center coordinates and period parameters. The state index refers to the retrieval marker composed of the operating wavelength state and polarization state, which distinguishes the target complex response of the same element in four states: 532 nm x-polarization, 532 nm y-polarization, 850 nm x-polarization, and 850 nm y-polarization. After all metasurface elements complete this binding process, the response constraint set solidifies the cross-band and cross-polarization optical targets into element-by-element records, providing a unique optical input for geometric inversion, and cross-band targets no longer rely on manual trade-offs.

[0036] Step 2 involves inverting the response constraint set to obtain the geometric target map and the geometric tolerance map. The geometric target map refers to the spatial parameter map of the recording unit number, geometric vector, and contour template. The geometric tolerance map refers to the spatial distribution map of the recording unit number and the allowable deviation of the parameters. The distinction between the two lies in the fact that the former provides the values ​​of the geometric parameters, while the latter provides the allowable deviation range for those values. The geometric vector is denoted as... ,in and This represents the two critical dimensions of the rectangular dielectric column. Indicates the direction of the outline.

[0037] Sub-step 2.1: According to the geometric vector of the discrete metasurface unit based on the manufacturing resolution limit, solve the complex transmission coefficient in the structure defined by the material level parameters to obtain the positive mapping from the geometric vector to the complex transmission coefficient.

[0038] Specifically, the manufacturing resolution limit refers to the minimum linewidth and dimensional adjustment step size given by the process, while the material layer parameters refer to the optical constants and layer thicknesses of each layer. A rigorous coupled-wave analysis model is used for the solution, which expands the periodic dielectric structure and electromagnetic field into spatial harmonics and outputs the complex transmission coefficient based on Maxwell's boundary conditions. and After discretization according to the size adjustment step size, the results of state-by-state simulation constitute a forward mapping, which refers to the lookup relationship between the geometric vector and the complex transmission coefficient.

[0039] Sub-step 2.2: The difference between the target complex response in the response constraint set and the forward mapping output is used to form the inversion objective function. The geometric vector is corrected within the value limited by the manufacturing resolution limit. The geometric vector that decreases with the inversion objective function is selected. The cell number, the geometric vector, and the contour template are written into the geometric target map.

[0040] Specifically, the inversion objective function is written as: ; in, Represents a geometric vector. Indicates the state index. This indicates a positive mapping onto a geometric vector. With state index The complex transmission coefficient of the output. This represents the target complex transmission coefficient in the constraint record. The correction employs a sequential quadratic programming method, which approximates the nonlinear objective function with a local quadratic function and updates the geometric vector within the values ​​defined by the manufacturing resolution limit. The profile template refers to the cross-sectional shape description of the rectangular medium column.

[0041] Sub-step 2.3: According to the size adjustment step size, the geometric vector is disturbed, and the change in complex transmission coefficient is obtained by positive mapping to form a sensitivity matrix. The preset complex response error threshold is divided by the sensitivity norm of the individual geometric parameter in the sensitivity matrix to obtain the parameter allowable deviation, and written into the geometric tolerance map.

[0042] Specifically, the sensitivity matrix refers to the set of first-order coefficients that cause changes in the complex transmission coefficient due to unit changes in each component of the geometric vector, given by a first-order Taylor sensitivity model. The preset complex response error threshold refers to the upper limit of the allowable complex response deviation for a single element, taken as 3%–4% of the target amplitude and 0.06–0.08 rad, an amplitude within the relatively flat range of cross-band focusing efficiency. The parameter allowable deviation refers to… , and Each element has its own permissible deviation range. The geometric tolerance chart converts optical tolerance into manufacturing tolerance, thus distinguishing between units with ample tolerance and those with stringent tolerance, and providing quantitative constraints on subsequent exposure adjustment intensity.

[0043] Step 3: Input the geometric target map and geometric tolerance map into the exposure transfer model to obtain the local sensitivity map and neighborhood coupling matrix. The exposure transfer model refers to the calculation process that converts the exposure dose into the final image size. It consists of three parts: projection energy diffusion calculation, photosensitive material state transition, and image transfer deviation mapping. Figure 1 As shown.

[0044] In sub-step 3.1, the exposure transfer model configures the exposure pattern according to the unit position and contour template in the geometric target image, performs projection energy diffusion calculation on the exposure dose of the metasurface unit, and obtains the substrate energy field.

[0045] Specifically, projected energy diffusion calculation refers to the calculation process of exposure energy spreading from the designed location to the surrounding area. It uses a double Gaussian point diffusion function model, with a narrow Gaussian kernel describing the main exposure area and a wide Gaussian kernel describing the neighboring diffusion area. The substrate energy field refers to the exposure energy per unit area obtained at each location on the substrate plane, calculated according to the following formula: ; in, Indicates position Substrate energy density at that location, Indicates the number of metasurface units. Indicates the unit number. Indicates the first Exposure dose per unit, Indicates the first The exposure area of ​​each unit, Indicates the first The center coordinates of each unit This indicates the energy transfer rate of the exposure system. This indicates the energy percentage of the narrow Gaussian kernel. and These represent the diffusion widths of the narrow Gaussian kernel and the wide Gaussian kernel, respectively.

[0046] Specifically, The initial value is 18–22 mJ / cm², which covers the effective exposure window of the photosensitive material. A value of 0.74 to 0.78 is used to allocate the main exposure energy to the contours within the cell. The size is 36–40 nm, which is on the same order of magnitude as the critical size of the unit, to control the internal contour of the control unit. Using 105–115 nm, close to the array period, reflects the energy penetration of adjacent units. The value is set to 0.90–0.94, consistent with the transmittance levels of the projection lens and mask. These values ​​are calibrated by fitting the exposure test pattern.

[0047] Sub-step 3.2 involves performing a photosensitive material state transition on the substrate energy field to obtain the latent image profile.

[0048] Specifically, the state transition of the photosensitive material is performed using the Dill photochemical model, which calculates the attenuation of photosensitive active components based on the substrate energy field. The latent image extraction threshold is set to 0.48–0.52 of the normalized attenuation, a range that falls within the main slope segment of the material response curve, where the distinction between exposed and unexposed areas is highest. The latent image profile refers to the contour lines where the attenuation equals the latent image extraction threshold, and their positions constitute the visible boundary after development.

[0049] Sub-step 3.3: Perform graphic transfer deviation mapping on the latent image profile to obtain the predicted geometric profile, apply positive and negative perturbations to the exposure dose of the disturbed unit, form a local sensitivity map based on the change intensity of the critical size of the disturbed unit, and form a neighborhood coupling matrix based on the change intensity of the critical size of the adjacent unit.

[0050] Specifically, the image transfer bias mapping refers to the process of converting the latent image profile into the processed structural profile. A morphological bias model is used, and the latent image boundary is moved along the normal direction by a process-calibrated bias distance to output the predicted geometric profile. The predicted geometric profile refers to the rectangular dielectric column boundary output by the exposure transfer model. The distinction between the predicted geometric profile and the target geometric parameters in the geometric target image lies in the fact that the former is calculated from the exposure dose, while the latter is calculated from optical inversion. The disturbed cell refers to the metasurface cell to which dose perturbation is applied, and the adjacent cell refers to the metasurface cell that shares an array edge with the disturbed cell. The perturbation ratio is taken as 2.0%–2.6%, which is within the local linear segment of dose to size and is higher than the numerical resolution threshold of the central finite difference method. The local sensitivity map records the response intensity of the disturbed cell's own critical size to dose changes. The neighborhood coupling matrix is... Line 1 Column element record The dose change of the first unit is directed towards the second unit. The dimensional changes transmitted by each unit. The non-uniformity of exposure energy transfer is thus transformed into quantitative data on a unit-by-unit and neighbor-by-neighbor pair basis, and differences that are difficult to observe under uniform exposure conditions are explicitly expressed.

[0051] Step 4: The partition index map and boundary coupling set are obtained by dividing the local sensitivity map and the neighborhood coupling matrix.

[0052] Sub-step 4.1: Using metasurface units as nodes, and the dimensionless value synthesized from the bidirectional transmission strength in the neighborhood coupling matrix, the sensitivity difference in the local sensitivity map, and the node center distance as connection weights, a weighted adjacency graph is constructed.

[0053] Specifically, the rectangular dielectric pillar located in the m-th row and n-th column constitutes node i, and the rectangular dielectric pillar to its right constitutes node k. The connection weight is calculated according to the following formula: ; in, This represents the connection weight between node i and node k. This represents the intensity of the change in the critical size of node k caused by a unit dose change at node i. This represents the strength transmitted from node k to node i. and Taken from local sensitivity map, Indicates a stable quantity. Indicates a scale for sensitivity differences. This represents the center distance between node i and node k. Indicates the spatial decay scale.

[0054] Specifically, all The median is denoted as the reference sensitivity. The reference sensitivity is set at 1.8% to 2.2%, which suppresses the sudden increase in weight caused by low-sensitivity nodes. Using 18% to 22% of the reference sensitivity, this range distinguishes nodes with similar exposure responses from nodes with significantly different responses. Taking 1.0 to 1.2 times the array period, this scale limits the contribution of far-distance units to the connection weight.

[0055] Sub-step 4.2: The sensitivity dispersion is the ratio of the standard deviation of sensitivity within the node set to the absolute value of the mean sensitivity. The node set of the weighted adjacency graph is recursively segmented according to the sensitivity dispersion threshold within the partition to obtain candidate partitions.

[0056] Specifically, the segmentation employs a normalized cut spectrum clustering algorithm. This algorithm establishes a normalized Laplacian operator based on connection weights and uses low-order eigenfunctions to separate weakly connected nodes. All nodes are first subjected to a binary split. Sets of nodes whose sensitivity dispersion is lower than the sensitivity dispersion threshold within a partition become candidate partitions. Sets with sensitivity dispersion higher than this threshold are further segmented using the same algorithm. The sensitivity dispersion threshold within a partition is set to 8%–10%, which matches the dose adjustment margin under a preset complex response error threshold. A candidate partition refers to a set of nodes that meets this threshold and has not yet undergone boundary correction.

[0057] Sub-step 4.3: The endpoints of an edge whose connection weight is higher than the inter-region coupling threshold and belongs to different candidate partitions are designated as the first boundary node and the second boundary node. The normalized cut cost of moving the first boundary node into the candidate partition of the second boundary node is compared with the normalized cut cost of moving the second boundary node into the candidate partition of the first boundary node. The move with the larger cost reduction is executed to obtain the exposure control partition, which is then written into the partition index graph and the boundary coupling set, as shown below. Figure 2 As shown.

[0058] Specifically, the Kernighan–Lin boundary exchange algorithm is used for the mobile operation. The inter-regional coupling threshold is set to 0.12–0.15, which is higher than the typical connection weight of the second-nearest unit within the array but lower than the connection weight of the strongly coupled boundary, thus stabilizing boundary identification. The exposure control zone refers to the set of nodes after boundary correction, and its number is written into the zone index map. The boundary coupling set refers to the data set recording the first boundary node number, the second boundary node number, and the connection weight of that edge. Units with similar dose responses are grouped into the same exposure control zone, while units with strong dose interference are explicitly registered at the boundary. Thus, the unified dose within a zone and the compensation across zones are each responsible for their respective functions.

[0059] Step 5: The zonal exposure parameters are obtained by solving the zonal index map, boundary coupling set, geometric target map, and geometric tolerance map. The zonal exposure parameters include the reference dose, boundary compensation amount, and scan connection position.

[0060] Sub-step 5.1: The exposure control partition is taken from the partition index map. The reference dose is the main dose inside the exposure control partition. The exposure control partitions connected by one record in the boundary coupling set are denoted as the first partition and the second partition. The boundary compensation amount is the dose correction value applied to the boundary band between the first partition and the second partition. The scan connection position is the boundary coordinate between the scan trajectory of the first partition and the scan trajectory of the second partition.

[0061] Understandably, the reference dose adjusts the critical size within the exposure control zone, the boundary compensation amount offsets the exposure energy diffused from the first zone to the second zone, and the scan connection position constrains the superposition width of the two scan tracks. The boundary midpoint refers to the midpoint of the line connecting the first boundary node and the second boundary node, and its coordinates serve as the initial value for the scan connection position.

[0062] Sub-step 5.2 converts the reference dose, boundary compensation amount, and scan connection position into iterative geometric parameters using the exposure transfer model. The objective function is then solved using the difference between the iterative geometric parameters and the target geometric parameters in the geometric target map, normalized by the geometric tolerance map; the ratio of the boundary compensation amount to the sum of the reference doses of the first and second partitions, weighted by the record connection weights; and the offset of the scan connection position relative to the midpoint of the boundary.

[0063] Specifically, iterative geometric parameters refer to the geometric parameters output by the exposure transfer model during the solution process. The distinction between iterative geometric parameters and predicted geometric contours lies in the fact that the former serves the iteration of this step, while the latter serves the residual comparison in step 6. The objective function is written as: ; in, This indicates solving for the objective function. This indicates the category of geometric parameters: 1 represents length, 2 represents width, and 3 represents contour direction. Represents the iterative geometric parameters. Taken from the geometric target image, Taken from the geometric tolerance diagram, Indicates the number of partition boundaries. Indicates the first Zone boundaries, Taken from the boundary coupling set, and Indicates the first The first and second partition numbers connected by the boundary line. and This indicates the baseline dose for both. Indicates the boundary compensation amount. Indicates the scan connection position. Indicates the coordinates of the midpoint of the boundary. Indicates the array period. , and This indicates setting parameters.

[0064] Specifically, With a weight of 0.18 to 0.22, this weight limits the boundary compensation to the range required for energy offset. The weight is set to 0.08 to 0.10, which constrains the offset of the scan trajectory. A value of 0.018 to 0.022 is used for weighted calculations on low-coupling boundaries. The upper limit of the amplitude is taken as 4% to 5% of the average value of the reference dose of the first zone and the reference dose of the second zone. relatively The maximum movement limit is 0.18. ~0.22 Both upper limits are consistent with the device's dose adjustment resolution and the model's local linear range.

[0065] Sub-step 5.3: Correct the reference dose, boundary compensation amount and scan connection position, and take the reference dose, boundary compensation amount and scan connection position within the range defined by the geometric tolerance map as the zonal exposure parameters.

[0066] Specifically, the correction employs a sequential quadratic programming algorithm. The baseline dose is initialized at 18–22 mJ / cm², the boundary compensation is initialized at zero, and the scan connection position is initialized at the midpoint of the boundary. Each round of correction calls the exposure transfer model to update the iterative geometric parameters, with the three types of parameters being jointly updated within the same objective function. In this way, the dose within a zone, the compensation across zones, and the trajectory connection are mutually balanced, and over-adjustment of a single parameter is suppressed by the latter two terms of the objective function.

[0067] Step 6: Input the partitioned exposure parameters into the exposure transfer model to obtain the predicted geometric profile. Obtain the two-domain residual from the predicted geometric profile, the geometric target map and the response constraint set. Use the two-domain residual to correct the partitioned exposure parameters to obtain the converged exposure parameters.

[0068] Sub-step 6.1: Extract the geometric parameters of the metasurface unit from the predicted geometric profile. The difference between the geometric parameters and the target geometric parameters in the geometric target map forms the geometric difference. Normalize the geometric difference using the geometric tolerance map to obtain the geometric residual.

[0069] Specifically, the length, width, and contour direction of node i are extracted from the predicted geometric contour. These are subtracted from the target geometric parameters in the geometric target map to obtain the geometric difference. The geometric difference is then divided by the allowable deviation of the parameters in the geometric tolerance map to obtain the geometric residual. A geometric residual greater than 1 indicates that the geometric parameter exceeds the manufacturing tolerance.

[0070] Sub-step 6.2: The complex transmission coefficient of the metasurface unit under the working wavelength and polarization states is reconstructed from the predicted geometric profile and denoted as the reconstructed complex response. The difference between the reconstructed complex response and the target complex response in the response constraint set forms the optical residual. The geometric residual and the optical residual are combined according to the preset weight to obtain the dual-domain residual.

[0071] Specifically, the reconstruction employs a rigorous coupled-wave analysis model. This model solves for Maxwell's boundary conditions based on the predicted geometric profile, outputting the amplitude and phase at four state indices for x-polarization and y-polarization at 532 nm and 850 nm. The reconstructed complex response refers to the complex transmission coefficient actually provided by the predicted structure; the distinction between the reconstructed complex response and the target complex response lies in the fact that the former is calculated from the predicted geometric profile, while the latter is taken from the constraint record. The dual-domain residual refers to the amount of error that simultaneously carries manufacturing deviations and optical response deviations.

[0072] Sub-step 6.3 uses the dual-domain residual to reverse-correct the reference dose, boundary compensation amount, and scan connection position along the local sensitivity map and the neighborhood coupling matrix to obtain the convergent exposure parameters.

[0073] Specifically, the correction employs the Levenberg-Marquardt algorithm, which uses damped least squares to determine the direction of residual descent. The local sensitivity map provides the propagation slope from the unit's own dose to the critical size, the neighborhood coupling matrix provides the slope from the dose change at node i to the contour at node k, and chain-reaction differentiation propagates the optical residual back to the geometric parameters, and the dual-domain residual back to the three types of zone exposure parameters. The inverse error of nodes within the exposure control zone corrects the reference dose, and the inverse error between the first and second boundary nodes corrects the boundary compensation amount according to the connection weights in the boundary coupling set. The error gradient of the scan overlay region moves the scan connection position. Parameters with normalized geometric residuals no greater than 1, amplitude errors no greater than 4%, and phase errors no greater than 0.10 rad are written into the convergent exposure parameters. Optical response, unit geometry, zone energy, and exposure execution form a closed control relationship within this loop, and the dimensional offset on the manufacturing side and the cross-band complex response deviation on the optical side decrease simultaneously in the same correction.

[0074] Step 7: Obtain the exposure execution sequence from the converged exposure parameters, the partition index map, and the boundary coupling set. Obtain the exposure execution deviation from the difference between the execution value of the exposure device and the exposure execution sequence. Use the exposure execution deviation to correct the exposure execution sequence of the unexposed partition to obtain the visible light near-infrared metasurface pattern.

[0075] Sub-step 7.1 involves arranging the exposure order using the partition index map and boundary coupling set, configuring energy commands and scan commands, and forming an exposure execution sequence.

[0076] Specifically, the maximum spanning tree algorithm is used for sequencing. This algorithm prioritizes connecting adjacent exposure control partitions with larger connection weights, and then performs a depth-first traversal of the spanning tree to output the exposure order from the first partition to the second partition. The controller reads the reference dose and scan endpoint of the first partition from the self-converging exposure parameters, and reads the boundary compensation amount and scan start point of the second partition. The energy command carries the reference dose and boundary compensation amount, and the scan command carries the scan connection position.

[0077] In sub-step 7.2, the energy sensor acquires the actual dose, and the position encoder acquires the actual scanning endpoint. The difference between the two and the command value in the exposure execution sequence constitutes the exposure execution deviation. The controller uses the exposure execution deviation to correct the energy command and scanning command of the unexposed zone.

[0078] Specifically, the correction employs a PID feedback algorithm, which generates a correction amount based on the current deviation, cumulative deviation, and deviation change rate. The dose deviation is written into the energy command of the unexposed zone, and the position deviation is written into the scan command of the unexposed zone. After all exposure control zones have completed exposure, the visible-near-infrared metasurface pattern is obtained through development and pattern transfer. The execution drift on the device side is successively absorbed between zones, and the deviation between the convergence result on the model side and the actual exposure result is suppressed. Cell critical size offset, boundary profile distortion, and cross-band complex response deviation are simultaneously narrowed across the entire metasurface.

[0079] This invention also discloses a visible light near-infrared metasurface unit zoned exposure control system, such as... Figure 3 As shown, the system includes a response constraint module, a geometry inversion module, an exposure transfer module, a partitioning module, a parameter solving module, a residual correction module, and an exposure execution module. The response constraint module binds the target complex response of the metasurface unit to the unit location according to the operating wavelength and polarization state, obtaining a response constraint set. The geometry inversion module inverts the response constraint set to obtain a geometric target map and a geometric tolerance map. The exposure transfer module inputs the geometric target map and geometric tolerance map into the exposure transfer model to obtain a local sensitivity map and a neighborhood coupling matrix. The partitioning module partitions the model using the local sensitivity map and neighborhood coupling matrix to obtain a partition index map and a boundary coupling set. The parameter solving module solves for the partition exposure parameters using the partition index map, boundary coupling set, geometric target map, and geometric tolerance map. The residual correction module inputs the partition exposure parameters into the exposure transfer model to obtain a predicted geometric profile, obtains a two-domain residual from the predicted geometric profile, geometric target map, and response constraint set, and uses the two-domain residual to correct the partition exposure parameters to obtain converged exposure parameters. The exposure execution module obtains the exposure execution sequence from the converged exposure parameters, the partition index map, and the boundary coupling set. The exposure execution deviation is obtained from the difference between the execution value of the exposure device and the exposure execution sequence. The exposure execution deviation is used to correct the exposure execution sequence of the unexposed partition to obtain the visible light near-infrared metasurface pattern.

[0080] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0081] The embodiments described above are merely illustrative of several implementations of the present invention, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of the present invention. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the present invention, and these modifications and improvements all fall within the scope of protection of the present invention. Therefore, the scope of protection of this patent should be determined by the appended claims.

Claims

1. A method for controlling the zoned exposure of visible and near-infrared metasurface units, characterized in that, include: The target complex response of the metasurface element is bound to the element location according to the working wavelength state and polarization state to obtain the response constraint set; The geometric target map and geometric tolerance map are obtained by inverting the response constraint set; the geometric target map and geometric tolerance map are input into the exposure transfer model to obtain the local sensitivity map and the neighborhood coupling matrix; The partitioned index map and boundary coupling set are obtained by dividing the local sensitivity map and the neighborhood coupling matrix; The zonal exposure parameters are obtained by solving the zonal index map, boundary coupling set, geometric target map, and geometric tolerance map. The zonal exposure parameters are input into the exposure transfer model to obtain the predicted geometric profile; the two-domain residuals are obtained from the predicted geometric profile, the geometric target map, and the response constraint set. The dual-domain residual is used to correct the partitioned exposure parameters to obtain the converged exposure parameters; The exposure execution sequence is obtained from the convergent exposure parameters, the partition index map, and the boundary coupling set; the exposure execution deviation is obtained from the difference between the execution value of the exposure device and the exposure execution sequence. The exposure execution sequence of the unexposed partition is corrected by the exposure execution deviation to obtain the visible light near-infrared metasurface pattern.

2. The method according to claim 1, characterized in that, The geometric target map and geometric tolerance map are obtained by inverting the response constraint set, specifically including: The geometric vectors of the discrete metasurface units are determined according to the manufacturing resolution limit. The complex transmission coefficient is solved in the structure defined by the material level parameters to obtain the forward mapping from the geometric vector to the complex transmission coefficient. The difference between the target complex response in the response constraint set and the output of the forward mapping is used to form the inversion objective function. The geometric vector is corrected within the value defined by the manufacturing resolution limit, and the geometric vector that decreases in the inversion objective function is selected. Write the element number, the geometric vector, and the contour template into the geometric target map; adjust the geometric vector according to the size adjustment step size, and obtain the change in the complex transmission coefficient by forward mapping to form a sensitivity matrix; divide the preset complex response error threshold by the sensitivity norm of the individual geometric parameters in the sensitivity matrix to obtain the parameter allowable deviation; write the element number and the parameter allowable deviation into the geometric tolerance map.

3. The method according to claim 1, characterized in that, The target complex response of the metasurface element is bound to the element location according to the operating wavelength state and polarization state, resulting in a response constraint set, which specifically includes: The target light field of the set focal plane is transformed to the spatial spectrum; the inverse propagation transfer function is applied to the spatial spectrum, and the metasurface plane complex light field under the working wavelength state is obtained by inverse transformation. The polarization basis vector is substituted into the model describing the vector transformation between the incident and emitted electric fields. The complex transmission coefficient under polarization state is extracted from the complex optical field of the metasurface plane. The maximum modulus normalization is performed on the complex transmission coefficient to obtain the target amplitude. The argument of the complex transmission coefficient is converted to a set phase period to obtain the target phase. The target amplitude and the target phase constitute the target complex response. The cell number is calculated according to the center coordinates and period parameters of the metasurface array. The state index is composed of the working wavelength state and the polarization state. The unit number, state index, and target complex response are combined to form a constraint record; the set of constraint records forms a response constraint set.

4. The method according to claim 1, characterized in that, The geometric target map and geometric tolerance map are input into the exposure transfer model to obtain the local sensitivity map and the neighborhood coupling matrix, specifically including: The exposure transfer model configures the exposure pattern according to the unit position and contour template in the geometric target image; the exposure transfer model performs projection energy diffusion calculation on the exposure dose of the metasurface unit to obtain the substrate energy field; and performs photosensitive material state transformation on the substrate energy field to obtain the latent image contour. A graphic transfer deviation mapping is performed on the latent image profile to obtain the predicted geometric profile; the metasurface unit to which dose perturbation is applied is denoted as the perturbed unit; the metasurface unit adjacent to the perturbed unit is denoted as the adjacent unit; positive and negative perturbations are applied to the exposure dose of the perturbed unit; A local sensitivity map is formed by the variation intensity of the critical size of the disturbed element relative to the positive and negative disturbances; a neighborhood coupling matrix is ​​formed by the variation intensity of the critical size of adjacent elements relative to the positive and negative disturbances.

5. The method according to claim 1, characterized in that, The partitioned index map and boundary coupling set are obtained by dividing the local sensitivity map and the neighborhood coupling matrix, specifically including: Using metasurface units as nodes, and dimensionless values ​​synthesized from the bidirectional transmission strength in the neighborhood coupling matrix, the sensitivity difference in the local sensitivity map, and the node center distance as connection weights, a weighted adjacency graph is constructed; the sensitivity dispersion is the ratio of the standard deviation of sensitivity within the node set to the absolute value of the mean sensitivity. The node set of the weighted adjacency graph is recursively segmented according to the sensitivity dispersion threshold within the partition to obtain candidate partitions; the endpoints of an edge whose connection weight is higher than the inter-partition coupling threshold and belongs to different candidate partitions are denoted as the first boundary node and the second boundary node. Compare the normalized cut cost of moving the first boundary node into the candidate partition to which the second boundary node belongs with the normalized cut cost of moving the second boundary node into the candidate partition to which the first boundary node belongs. Perform the move with the larger cost reduction to obtain the exposure control partition. Write the exposure control partition number to which the metasurface unit belongs into the partition index map; write the first boundary node number, the second boundary node number and the connection weight of the edge into the boundary coupling set.

6. The method according to claim 1, characterized in that, The zonal exposure parameters are obtained by solving the zonal index map, boundary coupling set, geometric target map, and geometric tolerance map, specifically including: The zonal exposure parameters include the reference dose, boundary compensation amount, and scan connection position; the exposure control zonal is taken from the zonal index map; the reference dose is the main dose within the exposure control zonal; the exposure control zonal connected by a record in the boundary coupling set is denoted as the first zonal and the second zonal; the boundary compensation amount is the dose correction value applied to the boundary band between the first zonal and the second zonal. The scan connection position is the boundary coordinate between the scan trajectory of the first zone and the scan trajectory of the second zone. The exposure transfer model converts the reference dose, boundary compensation amount, and scan connection position into iterative geometric parameters. The objective function is composed of the difference between the iterative geometric parameters and the target geometric parameters in the geometric target map, normalized by the geometric tolerance map; the ratio of the boundary compensation amount to the sum of the reference doses of the first zone and the second zone, weighted by the connection weight of the record; and the offset of the scan connection position relative to the midpoint of the boundary. The reference dose, boundary compensation amount, and scan connection position are corrected, and the reference dose, boundary compensation amount, and scan connection position that are within the range defined by the geometric tolerance map are taken as the zone exposure parameters.

7. The method according to claim 1, characterized in that, The two-domain residuals are obtained from the predicted geometric profile, the geometric target map, and the response constraint set, specifically including: The geometric parameters of the metasurface unit are extracted from the predicted geometric profile; the difference between the geometric parameters and the target geometric parameters in the geometric target map forms the geometric difference; the geometric difference is normalized using a geometric tolerance map to obtain the geometric residual; the complex transmission coefficient of the metasurface unit in the working wavelength state and polarization state is reconstructed from the predicted geometric profile, and is denoted as the reconstructed complex response; the difference between the reconstructed complex response and the target complex response in the response constraint set forms the optical residual; the geometric residual and the optical residual are combined according to a preset weight to obtain the dual-domain residual.

8. A visible-near-infrared metasurface unit zoned exposure control system, applied to the method described in any one of claims 1-7, characterized in that, include: The response constraint module binds the target complex response of the metasurface unit to the unit location according to the working wavelength state and polarization state, thus obtaining the response constraint set. The geometric inversion module inverts the response constraint set to obtain the geometric target map and the geometric tolerance map; The exposure transfer module inputs the geometric target map and geometric tolerance map into the exposure transfer model to obtain the local sensitivity map and the neighborhood coupling matrix. The partitioning module divides the local sensitivity map and the neighborhood coupling matrix to obtain a partition index map and a boundary coupling set; The parameter solving module obtains the zone exposure parameters from the zone index map, boundary coupling set, geometric target map, and geometric tolerance map. The residual correction module inputs the partitioned exposure parameters into the exposure transfer model to obtain the predicted geometric profile. The predicted geometric profile, the geometric target map and the response constraint set are used to obtain the two-domain residual. The two-domain residual is used to correct the partitioned exposure parameters to obtain the converged exposure parameters. The exposure execution module obtains the exposure execution sequence from the converged exposure parameters, the partition index map, and the boundary coupling set. The exposure execution deviation is obtained from the difference between the execution value of the exposure device and the exposure execution sequence. The exposure execution deviation is used to correct the exposure execution sequence of the unexposed partitions to obtain the visible light near-infrared metasurface pattern.

9. A terminal, comprising a processor and a storage medium; characterized in that: The storage medium is used to store instructions; The processor is configured to operate according to the instructions to perform the steps of the method according to any one of claims 1-7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When executed by a processor, the program implements the steps of the method according to any one of claims 1-7.

Images