Projection type structure photomask pattern manufacturability enhancing method and system

Abstract

The invention belongs to the technical field of projection light field regulation and control, and particularly relates to a method and a system for enhancing manufacturability of a projection type structure photomask pattern. Determining an initial mask pattern according to the target structured light field and the magnification of the projection system; calculating an initially generated structured light field by adopting a vector diffraction imaging model, calculating a numerical value of a target function, and judging whether the numerical value meets a preset threshold value or not; if the target function does not meet the preset threshold value, carrying out iterative optimization on the mask pattern by adopting an optimal gradient method until the preset threshold value of the target function is met, and outputting a continuous numerical value mask; pixel splitting or fusion is carried out on the continuous numerical mask according to the required mask manufacturing precision; local integral approximation operation is adopted, binarization processing is carried out on the mask, and the manufacturability of the mask is enhanced. According to the method, the projection type structure light field which is high in degree of freedom, high in efficiency, low in calculation power requirement and robust can be optimized, the mask pattern which can be manufactured can be obtained, and the manufacturing precision of the mask can be adjusted at will so as to adapt to different mask pattern preparation processes.

Description

Technical Field

[0001] The present invention belongs to the technical field of projection-generated structured light field control, and in particular relates to a method and system for enhancing the manufacturability of a projection-type structured light mask pattern. Background Art

[0002] The rapid development of modern science and technology has put forward more stringent requirements on the performance of optical systems, especially in cutting-edge fields such as microelectronics manufacturing, biomedical imaging, and high-precision measurement. The accuracy and complexity of optical components are facing new challenges. Take photolithography technology as an example. It is an indispensable core process in micro-nano manufacturing. It mainly realizes the precise transfer of patterns through the interaction between light and photosensitive materials (i.e., photoresists). This technology not only plays a key role in semiconductor manufacturing, but is also widely used in many important fields such as optical chips, optical communications, integrated photons, biosensors, and nanotechnology, becoming an important driving force for the development of modern high-tech industries. In the photolithography process, optical proximity correction technology (OPC) is a key link to ensure high-precision pattern transfer. As the semiconductor manufacturing process nodes continue to advance to smaller sizes, the resolution of the photolithography system gradually approaches the optical diffraction limit, and the influence of the optical proximity effect becomes more and more significant. This effect is mainly manifested in the deviation between the actual imaging and the designed pattern due to diffraction and interference of light waves during propagation, which is specifically manifested in problems such as rough edges, uneven line width, and pattern distortion. In order to overcome these challenges, structured light field control technology has become one of the core research directions of optical proximity correction.

[0003] In the projection system, the optimization of the light field intensity distribution directly determines the forming quality of the pattern on the photoresist. Traditional optical proximity correction technology mainly compensates for light field distortion by adjusting the geometric shape of the mask pattern, but its optimization goal is mainly focused on the amplitude control of the light field. However, it is often difficult to fully solve the problem of non-uniformity of light field intensity distribution by relying solely on amplitude control, especially in complex patterns or high-density integrated circuit designs. Local fluctuations in light field intensity may lead to underexposure or overexposure in key areas, thereby affecting the performance of the final device. Therefore, modern OPC technology has gradually introduced more sophisticated light field intensity distribution optimization strategies. By combining the light field propagation model and the inverse optimization algorithm, the transmittance distribution of the mask pattern is dynamically adjusted to achieve uniformity and accuracy of light intensity distribution on the photoresist surface.

[0004] In addition, the generation and regulation of structured light fields play an important role in the photolithography process. The intensity distribution of the structured light field not only affects the resolution and contrast of the pattern, but also determines the exposure effect of the photoresist. By optimizing the light field intensity distribution, the pattern distortion caused by the optical proximity effect can be effectively suppressed, and the stability and reliability of the photolithography process can be improved. For example, in deep subwavelength lithography, by introducing the light intensity distribution optimization algorithm, the local intensity of the light field can be accurately controlled in the mask design, thereby forming a high-fidelity micro-nano structure on the photoresist. This optimization method is not only applicable to traditional mask lithography, but can also be extended to advanced processes such as maskless lithography and extreme ultraviolet (EUV) lithography, providing technical support for the next generation of semiconductor manufacturing.

[0005] However, existing projection system light field optimization technologies usually introduce complex binary constraints in the optimization process, such as discrete optimization algorithms, regularization methods, and rule-based segmented optimization methods. These nonlinear complex calculation processes need to be coupled to each step of the iterative optimization process, which is a huge challenge to hardware computing power. Therefore, it is necessary to provide an improved projection structured light mask pattern manufacturability enhancement method to solve the above problems. Summary of the invention

[0006] The object of the present invention is to provide a method and system for enhancing the manufacturability of a projection structured light mask pattern. By performing binary restriction separation processing on mask pattern optimization and manufacturability enhancement, high-degree-of-freedom, high-efficiency, low-computing-power-requirement and robust projection structured light field optimization can be achieved and a manufacturable mask pattern can be obtained. Moreover, the manufacturing accuracy of the mask can be adjusted arbitrarily to adapt to different mask pattern preparation processes.

[0007] To achieve the above-mentioned object, the first aspect of the present invention provides a method for enhancing the manufacturability of a projection structured light mask pattern, comprising the following steps:

[0008] S1. According to the target light field I t , determine the initial mask pattern M ini ;

[0009] S2. Using the vector diffraction imaging model, the structured light field I generated by the projection system of the initial mask pattern is simulated. i ; and the Sigmoid function is used to calculate the generated structured light field profile Z i , enhancing the edge contrast of the structured light field;

[0010] S3, the target light field I t and the structured light field I generated by the simulation i Substitute the objective function, compare the calculated result of the objective function with the preset convergence threshold, and judge the structured light field I i Whether the requirements are met;

[0011] S4. If the objective function result is greater than the preset convergence threshold, the mask pattern M is optimized using the optimal gradient method. i The optimization update is performed, and then the structured light field generated by the updated mask pattern and its corresponding objective function result are calculated until the objective function meets the preset convergence threshold requirement or the number of optimizations reaches the upper limit, and the optimized mask pattern M with continuous values ​​is output. O ;

[0012] S5, using an interpolation method to optimize the mask pattern M O The pixels of the mask pattern are split or merged to adjust the pixel size of the mask pattern to obtain a mask pattern M with any desired pixel size. O-A ; Then the local integral approximation method is used to optimize the mask pattern M O-A Binarization restriction is performed to output a binary optimized mask pattern M with manufacturability O-L .

[0013] As a further preferred embodiment of the present invention, in step S1, the initial mask pattern M ini According to the target light field I t and the reduction ratio Mag of the projection system, Mag<1; I t is a matrix of size m*n, with pixel size p lf , M ini Is a matrix of size m*n, with pixel size P m , then P m =Mag*p lf .

[0014] As a further preferred embodiment of the present invention, in step S2, the generated structured light field I i The calculation method is:

[0015]

[0016]

[0017] Wherein, H represents the kernel function of the vector diffraction imaging model of the projection system in global coordinates; represents the convolution operation; E image represents the complex amplitude of the projection light field generated by simulation; M i represents the input mask, the subscript i represents the number of optimization iterations in step S3, and when i is 1, it represents the initial mask pattern M ini The generated initial structured light field; T g-g is the coordinate mapping relationship between the input field and the generated field, α ex , β ex and γ exrepresents the frequency domain coordinates on the projection plane; W is the window function of the projection system; α en and β en represents the frequency domain coordinate on the entrance plane, and NA represents the numerical aperture of the imaging lens.

[0018] As a further preferred embodiment of the present invention, in step S2, the Sigmoid function used is as follows:

[0019]

[0020] Z i Represents the light field profile corresponding to the generated light field, where a is the steepness parameter and b is the light intensity threshold for edge enhancement.

[0021] As a further preferred embodiment of the present invention, in step S3, the objective function F is as follows:

[0022]

[0023] ||||2 is the binary norm operation symbol. If the objective function F is less than the preset threshold δ, it means that the generated light field meets the requirements; if the objective function F is greater than the preset threshold δ, the mask pattern needs to be optimized.

[0024] As a further preferred embodiment of the present invention, in step S4, the intermediate variable matrix ω is introduced using trigonometric functions. i , derive the objective function F with respect to the intermediate variable matrix ω i The gradient expression of , and solve the optimal gradient, update the intermediate variable matrix ω according to the optimal gradient i+1 , then update the mask pattern M i+1 .

[0025] As a further preferred embodiment of the present invention, the intermediate variable matrix ω i and mask pattern M i The relationship is as follows:

[0026] ω i =cos -1 (2*M i -1)

[0027] Mask pattern M i The variable range is [0,1], and the intermediate variable matrix ω i M i The variable range is [-∞, +∞];

[0028] Objective function F and intermediate variable matrix ω i The gradient expression is as follows:

[0029]

[0030] In the iterative optimization process, according to the steepest descent method, the optimization speed along the negative gradient direction is optimally updated to update the intermediate variable matrix ω. i+1 :

[0031]

[0032] Among them, Δ represents the optimization step size; further, update the mask pattern:

[0033] M i+1 =(1+cos(ω i+1 )) / 2.

[0034] As a further preferred embodiment of the present invention, in step S5, the mask pattern M O-A Divide into several pixel units of size Λ*Λ, perform integration operation on the pixel values ​​in any unit, and use the integration result to approximate the integer value to determine the number of light-transmitting pixels in the unit. The principle expression of the local approximation method is as follows:

[0035]

[0036] Among them, Pixel_nu represents the number of light-transmitting pixels in the unit, Round represents the approximate rounding operation, j and k are the pixel index parameters in the pixel unit, and u and v represent the unit index parameters of the pixel unit.

[0037] A second aspect of the present invention provides a projection structured light mask pattern manufacturability enhancement system, comprising:

[0038] A calculation module is used to determine an initial mask pattern according to a target light field; and ini Calculate the structured light field I generated after passing through the projection system i ;

[0039] The judgment module is used to convert the target light field I t The parameters and structured light field I i Substituting the objective function F into the target function, determining whether the target function F meets the pre-convergence threshold requirement;

[0040] Optimization module for mask pattern M i If the objective function F does not meet the preset convergence threshold requirement, the optimal gradient method is used to optimize the mask pattern M i The pixel values ​​are optimized and updated until the objective function F meets the preset convergence threshold requirement or the number of optimizations reaches the preset number, and the optimized mask pattern M with continuous values ​​is output. O ;

[0041] Binarization module is used to convert the mask pixel size P mAdjust to any desired size to obtain a mask pattern M with any desired pixel size O-A , and the optimized mask pattern M with continuous values O-A Converted into a mask pattern M with binary values O-L .

[0042] A third aspect of the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein when the computer program is executed by a processor, the method for enhancing the manufacturability of a projection structured light mask pattern as described in any one of the above items is implemented.

[0043] In general, the above technical solution conceived by the present invention has the following technical advantages compared with the prior art:

[0044] 1. The method for enhancing the manufacturability of projection structured light mask patterns provided by the present invention simplifies the complex pattern optimization process and reduces the demand for computing resources by step-by-step processing of the binary restriction process of mask pattern optimization and manufacturability enhancement. At the same time, it can achieve high degree of freedom, high efficiency, low computing power demand and robust projection structured light field optimization, and obtain a manufacturable mask pattern, and the manufacturing accuracy of the mask can be arbitrarily adjusted to adapt to different mask pattern preparation processes.

[0045] 2. In the mask binarization restriction process of the present invention, the mask pattern is divided into a number of pixel units, and a high-fidelity mask binarization process can be achieved by using simple linear operations.

[0046] 3. The present invention introduces the local integral approximation method into the projection structured light field optimization process for the first time, so that the mask pattern optimization process has a high degree of optimization freedom; by using interpolation operations, the pixels of the mask pattern are split and merged to obtain any desired pixel size, thereby meeting the accuracy requirements of different mask manufacturing processes. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] Figure 1 It is a schematic diagram of a flow chart of a method for enhancing the manufacturability of a projection structured light mask pattern provided by the present invention.

[0048] Figure 2 It is a schematic diagram of the initial input / target structured light field selected in the specific example provided by the present invention.

[0049] Figure 3 It is a schematic diagram of generating a structured light field using an initial input light field in a specific example provided by the present invention.

[0050] Figure 4 It is a schematic diagram of the optimized continuous numerical mask pattern in the specific example provided by the present invention.

[0051] Figure 5It is a schematic diagram of a structured light field generated by a continuous numerical mask pattern in a specific example provided by the present invention.

[0052] Figure 6 It is a schematic diagram of the binarization process of the mask pattern in the specific example provided by the present invention.

[0053] Figure 7 It is a schematic diagram of a structured light field generated by a binary mask pattern in a specific example provided by the present invention.

[0054] Figure 8 It is a schematic diagram of the intensity difference between the light field generated by the continuous numerical mask and the binary mask in the specific example provided by the present invention. DETAILED DESCRIPTION

[0055] In order to make the purpose, technical scheme and advantages of the present invention clearer, the present invention is further described in detail below in conjunction with the embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention. In addition, the technical features involved in the various embodiments of the present invention described below can be combined with each other as long as they do not conflict with each other.

[0056] See also Figure 1 The first embodiment provides a method for enhancing the manufacturability of a projection structured light mask pattern, the method comprising:

[0057] According to the target light field I t , determine the initial mask pattern M ini ,like Figure 2 As shown;

[0058] Using the vector diffraction imaging model, the structured light field generated by the mask pattern through the projection system is simulated and calculated. i ,like Figure 3 As shown;

[0059] And use the Sigmoid function to calculate the light field profile Z i , Enhanced simulation to generate structured light field I i edge contrast;

[0060] Project the target light field I t and the structured light field I generated by the simulation i Substitute the objective function F, compare the calculated result of the objective function with the preset convergence threshold, and judge the generated structured light field I i Whether the requirements are met;

[0061] If the objective function result is greater than the preset convergence threshold, the optimal gradient method is used to adjust the mask pattern M iPerform optimization and update, then calculate the structured light field generated by the updated mask pattern and its corresponding objective function result, until the objective function F meets the preset convergence threshold requirement or the number of optimizations reaches the upper limit, and output the optimized mask pattern M with continuous values O ;

[0062] Sampling interpolation calculation for optimizing mask pattern M O The pixels of the mask pattern are split or merged to adjust the pixel size of the mask pattern to obtain a mask pattern M with any desired pixel size. O-A ; Then the local integral approximation method is used to optimize the mask pattern M O-A Binarization is performed to limit the output of a binary optimized mask pattern M with excellent manufacturability O-L .

[0063] In this way, a step-by-step mask pattern optimization and binarization constraint method is adopted. By dividing the mask pattern into pixel units and performing local integral approximation processing, a high-fidelity binary pattern can be obtained using simple linear operations, providing technical support for the further development of projection light field control technology, and solving the problem that the existing structured light field optimization technology cannot achieve the optimal design of binary mask patterns with high efficiency and high fidelity, and requires complex calculation models and high computing power consumption.

[0064] Specifically, according to the target light field I t The initial mask pattern M is determined by the reduction factor Mag (Mag<1) of the projection system. ini , I t is a matrix of size m*n, with pixel size p lf , M ini is a matrix of size m*n, pixel size P m , where P m =Mag*p lf .

[0065] Adopting the vector diffraction imaging model suitable for high NA (numerical aperture) and polarized light imaging, the spatial distribution of physically generated light fields is accurately simulated and calculated. i , the subscript indicates the number of iterations, i = 1, 2, 3, ..., i = 1 indicates the initial mask pattern M ini The generated initial structured light field;

[0066] The structured light field I generated by the simulation i The calculation method of the strength can be expressed as:

[0067]

[0068]

[0069] Wherein, H represents the kernel function of the vector diffraction imaging model of the projection system in global coordinates; represents the convolution operation; E image represents the complex amplitude of the projection light field generated by simulation; M i represents the input mask, the subscript i represents the number of iterations, and when i is 1, it represents the result corresponding to the initial mask; T g-g is the coordinate mapping relationship between the input field and the generated field, α ex , β ex and γ ex represents the frequency domain coordinates on the projection plane; W is the window function of the projection system; α en and β en represents the frequency domain coordinate on the entrance plane, and NA represents the numerical aperture of the imaging lens.

[0070] The Sigmoid function is used to calculate the outline of the light field to enhance the subsequent optimization effect of the light field edge. The Sigmoid function used can be expressed as:

[0071]

[0072] Z i Represents the light field profile corresponding to the generated light field, where a is the steepness parameter and b is the light intensity threshold for edge enhancement.

[0073] The objective function F can be written as:

[0074] F=||Z i -I t ||2 2

[0075] || ||2 is the binary norm operation symbol. If the objective function F is less than the preset threshold δ, it means that the generated light field meets the requirements; if the objective function F is greater than the preset threshold δ, the mask pattern needs to be optimized.

[0076] Using the optimal gradient method, the mask pattern M i To optimize, use trigonometric functions to introduce the intermediate variable matrix ω i , the relationship between the two is as follows:

[0077] ω i =cos -1 (2*M i -1)

[0078] Mask pattern M i The variable range is [0,1], and the intermediate variable matrix ω i M iThe variable range is [-∞, +∞] (i.e., negative infinity to positive infinity), and the light field transforms the discontinuous optimization problem of mask values ​​with a finite variable range into a continuous optimization problem with an infinite variable interval.

[0079] Derivation of the objective function F with respect to the intermediate variable matrix ω i The gradient expression of , and solve the optimal gradient, update the intermediate variable matrix ω according to the optimal gradient i+1 , then update the mask pattern M i+1 .

[0080]

[0081] According to the chain derivation rule, the gradient expression of the objective function F with respect to the intermediate variable ω can be obtained as follows:

[0082]

[0083] In the iterative optimization process, according to the steepest descent method, the optimization speed along the negative gradient direction is the best, and the intermediate variables are updated:

[0084]

[0085] Where Δ represents the optimization step size. Further, update the mask pattern:

[0086] M i+1 =(1+cos(ω i+1 )) / 2.

[0087] Calculate the spatial image corresponding to the updated mask pattern and evaluate whether the value of the objective function F reaches the preset convergence threshold. If it has converged, the iterative optimization stops. Otherwise, repeat the optimization process until the convergence condition is reached or the number of iterations reaches the upper limit, and output the optimized mask pattern M with continuous values. O ,like Figure 4 As shown. Optimize the mask pattern M O The generated optimized structured light field is as follows Figure 5 shown.

[0088] The numerical interpolation method is used to optimize the mask pattern M O The pixel size is reconstructed, including pixel splitting and fusion, to obtain a mask pattern M with any desired pixel size. O-A ;

[0089] The local approximation method is used to continuously optimize the mask pattern M O-A Binarization is performed and the mask pattern M O-ADivide into several pixel units of size Λ*Λ, perform integration operation on the pixel values ​​in any unit, and use the integral result to approximate the integer value to determine the number of transparent (value 1) pixels in the unit. The principle expression of the local approximation method is as follows:

[0090]

[0091] Among them, Pixel_nu represents the number of light-transmitting pixels in the unit, Round represents the approximate rounding operation, j and k are the pixel index parameters in the pixel unit, and u and v represent the unit index parameters of the pixel unit; the local integral approximation method outputs the binary optimized mask pattern M O-L .

[0092] In this example, Λ is taken as 3, and the process of the local integral approximation method is as follows: Figure 6 shown.

[0093] Calculate the binary mask M using the vector diffraction imaging model O-L The generated optimized structured light field, such as Figure 7 As shown. Continuous value mask M O-A With the binary mask M O-L The intensity distribution of the generated structured light fields is basically the same, and the difference results are as follows: Figure 8 shown.

[0094] According to another aspect of the present invention, there is provided a system for enhancing the manufacturability of a projection structured light mask pattern, comprising:

[0095] A calculation module is used to determine an initial mask pattern according to a target light field; and ini Calculate the structured light field I generated after passing through the projection system i ;

[0096] The judgment module is used to convert the target light field I t The parameters and structured light field I i Substituting the objective function F into the target function, determining whether the target function F meets the pre-convergence threshold requirement;

[0097] Optimization module for mask pattern M i If the objective function F does not meet the preset convergence threshold requirement, the optimal gradient method is used to optimize the mask pattern M i The pixel values ​​are optimized and updated until the objective function F meets the preset convergence threshold requirement or the number of optimizations reaches the preset number, and the optimized mask pattern M with continuous values ​​is output. O ;

[0098] Binarization module is used to convert the mask pixel size P mAdjust to any desired size to obtain a mask pattern M with any desired pixel size O-A , and the optimized mask pattern M with continuous values O-A Converted into a mask pattern M with binary values O-L .

[0099] In summary, the present invention proposes a method for enhancing the manufacturability of a projection structured light mask pattern, which separates the mask pattern optimization and the binarization restriction, and introduces the local integral approximation method into the projection structured light field optimization process for the first time, so that the mask pattern optimization process has a high degree of optimization freedom, and the binarization restriction process used to enhance the manufacturability of the mask pattern has the characteristics of high fidelity and high efficiency. Therefore, the proposed method can achieve fast and accurate mask pattern optimization under the premise of low computing power consumption.

[0100] It will be easily understood by those skilled in the art that the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for enhancing the manufacturability of a projection structured light mask pattern, characterized in that: The following steps are involved: S1. According to the target light field I t , determine the initial mask pattern M ini ; S2. Using the vector diffraction imaging model, the structured light field I generated by the projection system of the initial mask pattern is simulated. i ; S3, the target light field I t and the structured light field I generated by the simulation i Substitute the objective function, compare the calculated result of the objective function with the preset convergence threshold, and judge the structured light field I i Whether the requirements are met; S4. If the objective function result is greater than the preset convergence threshold, the mask pattern M is optimized using the optimal gradient method. i The optimization update is performed, and then the structured light field generated by the updated mask pattern and its corresponding objective function result are calculated until the objective function meets the preset convergence threshold requirement or the number of optimizations reaches the upper limit, and the optimized mask pattern M with continuous values ​​is output. O ; S5, using an interpolation method to optimize the mask pattern M O The pixels of the image are split or merged to obtain a mask pattern M with any desired pixel size. O-A ; Then the local integral approximation method is used to optimize the mask pattern M O-A Binarization restriction is performed to output a binary optimized mask pattern M with manufacturability O-L .

2. The method for enhancing the manufacturability of projection structured light mask patterns according to claim 1, characterized in that: In step S1, the initial mask pattern M ini According to the target light field I t and the reduction ratio Mag of the projection system, Mag<1; I t is a matrix of size m*n, with pixel size p lf , M ini Is a matrix of size m*n, with pixel size P m , then P m =Mag*p lf .

3. The method for enhancing the manufacturability of projection structured light mask patterns according to claim 1, characterized in that: In step S2, the generated structured light field I i The calculation method is: Wherein, H represents the kernel function of the vector diffraction imaging model of the projection system in global coordinates; represents the convolution operation; E image represents the complex amplitude of the projection light field generated by simulation; M i represents the input mask, the subscript i represents the number of optimization iterations in step S3, and when i is 1, it represents the initial mask pattern M ini The generated initial structured light field; T g-g is the coordinate mapping relationship between the input field and the generated field, α ex , β ex and γ ex represents the frequency domain coordinates on the projection plane; W is the window function of the projection system; α en and β en represents the frequency domain coordinate on the entrance plane, and NA represents the numerical aperture of the imaging lens.

4. The method for enhancing the manufacturability of projection structured light mask patterns according to claim 1, characterized in that: In step S2, the Sigmoid function is also used to calculate the generated structured light field profile Z i , to enhance the structured light field I i The edge contrast of; the Sigmoid function used is as follows: Z i Represents the light field profile corresponding to the generated light field, where a is the steepness parameter and b is the light intensity threshold for edge enhancement.

5. The method for enhancing the manufacturability of projection structured light mask patterns according to claim 1, characterized in that: In step S3, the objective function F is as follows: ||||2 is the binary norm operation symbol. If the objective function F is less than the preset threshold δ, it means that the generated light field meets the requirements; if the objective function F is greater than the preset threshold δ, the mask pattern needs to be optimized.

6. The method for enhancing the manufacturability of projection structured light mask patterns according to claim 1, characterized in that: In step S4, the intermediate variable matrix ω is introduced using trigonometric functions i , derive the objective function F with respect to the intermediate variable matrix ω i The gradient expression of , and solve the optimal gradient, update the intermediate variable matrix ω according to the optimal gradient i+1 , then update the mask pattern M i+1 .

7. The method for enhancing the manufacturability of projection structured light mask patterns according to claim 6, characterized in that: Intermediate variable matrix ω i and mask pattern M i The relationship is as follows: oh i =cos -1 (2*M i -1) Mask pattern M i The variable range is [0,1], and the intermediate variable matrix ω i M i The variable range is [-∞, +∞]; Objective function F and intermediate variable matrix ω i The gradient expression is as follows: In the iterative optimization process, according to the steepest descent method, the optimization speed along the negative gradient direction is optimally updated to update the intermediate variable matrix ω. i+1 : Among them, Δ represents the optimization step size; finally update the mask pattern: M i+1 =(1+cos(ω) i+1 )) / 2。 8. The method for enhancing the manufacturability of projection structured light mask patterns according to claim 1, characterized in that: In step S5, the mask pattern M O-A Divide into several pixel units of size Λ*Λ, perform integration operation on the pixel values ​​in any unit, and use the integration result to approximate the integer value to determine the number of light-transmitting pixels in the unit. The principle expression of the local approximation method is as follows: Among them, Pixel_nu represents the number of light-transmitting pixels in the unit, Round represents the approximate rounding operation, j and k are the pixel index parameters in the pixel unit, and u and v represent the unit index parameters of the pixel unit.

9. A projection structured light mask pattern manufacturability enhancement system, characterized in that: include: A calculation module, used for determining an initial mask pattern according to a target light field; And according to the initial mask M ini Calculate the structured light field I generated after passing through the projection system i ; The judgment module is used to convert the target light field I t The parameters and structured light field I i Substituting the objective function F into the target function, determining whether the target function F meets the pre-convergence threshold requirement; Optimization module for mask pattern M i If the objective function F does not meet the preset convergence threshold requirement, the optimal gradient method is used to optimize the mask pattern M i The pixel values ​​are optimized and updated until the objective function F meets the preset convergence threshold requirement or the number of optimizations reaches the preset number, and the optimized mask pattern M with continuous values ​​is output. O ; Binarization module is used to convert the mask pixel size P m Adjust to any desired size to obtain a mask pattern M with any desired pixel size O-A , and the optimized mask pattern M with continuous values O-A Converted into a mask pattern M with binary values O-L .

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the method for enhancing the manufacturability of a projection structured light mask pattern according to any one of claims 1 to 8 is implemented.

Images