Method for calibrating multi-space sampling rate response matrix of deformable mirror

Abstract

The present disclosure relates to the technical field of light field wavefront sensing, and particularly relates to a method for calibrating a multi-space sampling rate response matrix of a deformable mirror, which adopts a checkerboard algorithm to obtain a response matrix of the highest space sampling rate; receives a user input of a wavefront target space sampling rate, calculates a down-sampling window size of the response matrix of the highest space sampling rate; down-samples the response matrix of the highest space sampling rate to generate a slope vector of the target space sampling rate, and performs a slope consistency evaluation within the down-sampling window, judges whether the wavefront target space sampling rate is suitable for the current wavefront distribution according to the evaluation result; and splices the slope vectors of all the wavefront target space sampling rates to generate a response matrix of the target space sampling rate. Only a single execution of the calibration of the highest space sampling rate is required, and the response matrix of the target space sampling rate can be generated as required, thereby reducing the calibration workload and improving the system efficiency.

Description

Technical Field

[0001] This disclosure relates to the field of optical wavefront sensing technology, specifically to a method for calibrating the multi-space sampling rate response matrix of a deformable mirror. Background Technology

[0002] Adaptive optics systems have wide applications in fields such as astronomical observation, free-space optical communication, and directed energy. Their core component, the wavefront sensor, is used to measure the wavefront distortion information of the incident light field in real time, providing feedback for the wavefront correction of the system.

[0003] Traditional optical wavefront sensors (such as the Shack-Hartmann sensor) have limitations in terms of dynamic range. In recent years, optical wavefront sensors, as an emerging wavefront detection method, can simultaneously acquire spatial and angular information of the optical field, have a larger dynamic measurement range and stronger wavefront reconstruction capabilities, and therefore show significant advantages in applications such as free-space optical communication in complex atmospheric turbulence environments and directed energy applications.

[0004] In adaptive optics systems, high-precision wavefront correction typically requires calibration of deformable mirrors to establish a response matrix. In the calibration of adaptive optics systems incorporating wavefront sensors, response matrices calibrated with high wavefront spatial sampling rates offer advantages such as high accuracy and good fitting performance. However, in practical wavefront correction, they also suffer from drawbacks including poor noise immunity, long computation time, and limited closed-loop bandwidth. Therefore, it is necessary to dynamically adjust different wavefront spatial sampling rates based on specific circumstances.

[0005] To meet these requirements, existing push-pull or pattern-based methods require multiple calibrations at different wavefront spatial sampling rates to generate response matrices corresponding to these rates. These methods are not only cumbersome and time-consuming, but also produce significant slope errors in areas of strong turbulence or low light intensity, leading to a decrease in the overall robustness of the system. Summary of the Invention

[0006] To address the problems existing in the prior art, the purpose of this disclosure is to provide a method for calibrating the multi-space sampling rate response matrix of a deformable mirror. This method solves the problems of traditional adaptive optics systems containing light field sensors requiring multiple calibrations for different target space sampling rates, resulting in long calibration times and large slope errors in areas with strong turbulence or low light intensity.

[0007] To achieve the above objectives, a method for calibrating the multi-space sampling rate response matrix of a deformable mirror is provided according to an embodiment of this disclosure. The method includes:

[0008] The response matrix with the highest spatial sampling rate is obtained by using a checkerboard algorithm on the voltage calibration image acquired by the optical wavefront sensor.

[0009] Receive the wavefront target spatial sampling rate input by the user, and calculate the downsampling window size of the response matrix with the highest spatial sampling rate;

[0010] The response matrix of the highest spatial sampling rate is downsampled to generate the slope vector of the target spatial sampling rate, and the slope consistency within the downsampling window is evaluated. Based on the evaluation results, it is determined whether the target spatial sampling rate of the wavefront is suitable for the current wavefront distribution. If it is not suitable, the target spatial sampling rate of the wavefront is increased.

[0011] If the conditions are met, the slope vectors of all wavefront target space sampling rates are concatenated to generate the target space sampling rate response matrix.

[0012] Furthermore, the response matrix with the highest spatial sampling rate is obtained using the checkerboard algorithm, including:

[0013] Each calibration image is segmented based on the sub-aperture of the microlens array to obtain a sub-aperture image;

[0014] Within the central rectangular region of each sub-aperture image, an inscribed circle binary mask is generated to obtain the sub-aperture image after applying the mask;

[0015] Find the maximum grayscale value in the set of pixel values ​​corresponding to the same local coordinates in each sub-aperture image;

[0016] Record the maximum grayscale value and the center coordinates of the sub-aperture image where the maximum grayscale value is located;

[0017] By traversing all local coordinates, the corresponding maximum gray values ​​are used to form a light intensity distribution matrix, and the center coordinates of the sub-aperture image where the maximum gray value is located are used to form a sub-aperture coordinate matrix.

[0018] Convert the sub-aperture coordinate matrix into a wavefront slope matrix;

[0019] The wavefront slope matrices of all the calibration images are concatenated to form the response matrix with the highest spatial sampling rate.

[0020] Furthermore, the slope consistency assessment within the downsampling window includes:

[0021] Divide the sub-aperture coordinate matrix into non-overlapping downsampling windows;

[0022] Calculate the local slope variation coefficients in the x and y directions within each downsampling window;

[0023] The local slope variation coefficients in the x and y directions are compared with the variation threshold. If there is a local slope variation coefficient greater than the variation threshold, it is determined that the wavefront target space sampling rate is not suitable for the current wavefront distribution.

[0024] Further, the calculation of the local slope variation coefficients in the x and y directions within each downsampling window includes: calculating the variance of the slope values ​​in the x and y directions within the downsampling window as the local slope variation coefficients in the x and y directions within the downsampling window, respectively.

[0025] Furthermore, the response matrix at the highest spatial sampling rate is downsampled to generate the slope vector of the target spatial sampling rate, including:

[0026] Based on the uniformity of light intensity distribution, the average value of the slope in the x-direction and the average value of the slope in the y-direction of the target space sampling rate for each calibration image are calculated by selecting either the uniform averaging mode or the light intensity weighted averaging mode, and then concatenated into a long vector. The corresponding long vectors of all calibration images are then concatenated to form the target space sampling rate response matrix.

[0027] Furthermore, the calibration images are obtained by using a region calibration method or a mode calibration method, applying a unit voltage or a Zernike mode voltage to each deformable mirror actuator individually, and recording the calibration images of the optical wavefront sensor for the voltage one by one.

[0028] Compared with the prior art, the advantages of this disclosure are:

[0029] This disclosure requires only a single execution of the highest spatial sampling rate calibration to generate the target spatial sampling rate response matrix as required, reducing calibration workload and improving system efficiency. Verification shows that the cosine similarity between the fitted target spatial sampling rate response matrix and the truly calibrated target spatial sampling rate response matrix is ​​over 90%. Attached Figure Description

[0030] To more clearly illustrate the technical solutions in the embodiments of this disclosure, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this disclosure. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0031] Figure 1 A flowchart of the method provided for implementation of this disclosure;

[0032] Figure 2 A flowchart of a method for obtaining the response matrix with the highest spatial sampling rate using a checkerboard algorithm, provided for embodiments of this disclosure;

[0033] Figure 3 A flowchart illustrating the method for evaluating slope consistency within a downsampling window provided for implementation of this disclosure. Detailed Implementation

[0034] To make the objectives, technical solutions, and advantages of the embodiments of this disclosure clearer, the technical solutions of the embodiments of this disclosure will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this disclosure. Based on the embodiments of this disclosure, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this disclosure.

[0035] The embodiments of this disclosure will be further described in detail below with reference to the accompanying drawings.

[0036] See Figure 1 The method shown is for calibrating the multi-space sampling rate response matrix of a deformable mirror, the method comprising:

[0037] S1 uses a checkerboard algorithm to obtain the response matrix with the highest spatial sampling rate from the voltage calibration image acquired by the optical wavefront sensor.

[0038] The calibration images were obtained by using either the region calibration method or the mode calibration method. A unit voltage or Zernike mode voltage was applied to each deformable mirror actuator individually, and the calibration images of the optical wavefront sensor for the voltage were recorded one by one.

[0039] For the region calibration method, a unit voltage is applied to each deformable mirror actuator individually, and the calibration image of the light field sensor for the voltage is recorded for each actuator. For the mode calibration method, a unit Zernike mode voltage is applied to each deformable mirror individually, and the calibration image of the light field sensor for the voltage is recorded for each mirror.

[0040] Among them, the regional calibration method refers to a calibration method in adaptive optics systems that directly uses the measured value of the local wavefront slope output by the wavefront sensor as the basic quantity to establish a linear correspondence between it and the driving signals of each actuator of the deformable mirror.

[0041] The mode calibration method refers to a calibration method in adaptive optics systems that first reconstructs the slope data measured by the wavefront sensor into coefficients of a selected set of orthogonal or quasi-orthogonal modes (such as Zernike polynomials), and then establishes a linear correspondence between these mode coefficients and the driving signals of each actuator of the deformable mirror.

[0042] S2 receives the wavefront target spatial sampling rate input by the user and calculates the downsampling window size of the response matrix of the highest spatial sampling rate;

[0043] S3 downsamples the response matrix of the highest spatial sampling rate to generate the slope vector of the target spatial sampling rate, and performs slope consistency evaluation within the downsampling window. Based on the evaluation results, it is determined whether the target spatial sampling rate of the wavefront is suitable for the current wavefront distribution. If it is not suitable, the target spatial sampling rate of the wavefront is increased.

[0044] If S4 is satisfied, then the slope vectors of all wavefront target space sampling rates are concatenated to generate the target space sampling rate response matrix.

[0045] In one embodiment, see Figure 2 As shown, the response matrix obtained using the checkerboard algorithm with the highest spatial sampling rate includes:

[0046] S101 divides each calibration image into sub-aperture images based on the sub-aperture of the microlens array. A sub-aperture of the microlens array refers to a local measurement unit comprised of each independent microlens in the microlens array (MLA) and its corresponding incident pupil sampling area. The microlens array used includes... Each microlens unit is arranged in a square, and the sub-aperture image size corresponding to each microlens is [size missing]. Pixels, therefore, each calibrated image segmentation yields Each sub-aperture image is a... Pixel size, recording the (m,n)th sub-aperture image as ,in , To establish a Cartesian coordinate system with the central sub-aperture of the microlens array as the origin, the sub-aperture row and column indexes are... Let be the pixel index within the sub-aperture image, i.e., the local coordinates. Then, the pixel correspondence between the calibration image and the sub-aperture image can be expressed as:

[0047] ,

[0048] in, To calibrate the image, , , ,and All represent the segmented sub-aperture images. Represents an integer;

[0049] S102 generates an inscribed circle binary mask within the central rectangular region of each sub-aperture image to obtain the sub-aperture image after applying the mask;

[0050] Center of each sub-aperture image Generate a binary mask of the inscribed circle within the rectangular region. Pixels within the mask participate in grayscale summation, while pixels outside the mask are invalid pixels. The diameter of the region mapped by the pupil in the sub-aperture image is given by the microlens array. Each sub-aperture of the microlens array corresponds precisely to a microlens spacing on the system pupil plane. The local pupil area. Microlens spacing: Microlens spacing refers to the center-to-center distance between two adjacent microlenses in a microlens array. , The image after masking is labeled as square after image segmentation. An inscribed circle is added to each segmented square image. The pixels inside the circle are valid pixels used in the calculation, while the pixels at the four corners are invalid values. (not a number) indicates.

[0051] The sub-aperture image after applying the mask is as follows: , This is the sub-aperture image after applying a mask.

[0052] Because of all The mask shape, size, and relative position of each sub-aperture image are exactly the same; therefore, in each calibration image, the same local coordinates of each sub-aperture image are identical. The corresponding place A set of pixel values :

[0053] .

[0054] S103 searches for the maximum grayscale value in the set of pixel values ​​corresponding to the same local coordinates in each sub-aperture image: , Indicates the maximum grayscale value;

[0055] S104 records the maximum grayscale value and the center coordinates of the sub-aperture image where the maximum grayscale value is located. ; This indicates the center coordinates of the sub-aperture image.

[0056] The maximum grayscale value and the center coordinates of the sub-aperture image containing the maximum grayscale value are both within the sub-aperture row and column index set: , It is an odd number;

[0057] S105 traverses all local coordinates and forms a light intensity distribution matrix by the corresponding maximum gray values. The center coordinates of the sub-aperture image where the maximum gray value is located form the sub-aperture coordinate matrix.

[0058] Traverse all local coordinates within the mask Then, the maximum grayscale value That is, to constitute a picture The size of the light intensity distribution matrix, i.e. The center coordinates of the sub-aperture image where the maximum grayscale value is located It constitutes The sub-aperture coordinate matrix corresponding to each pixel in the image, and the two represent the local coordinates respectively. Light intensity information in the calibration image and and Slope information of the direction.

[0059] S106 converts the sub-aperture coordinate matrix into a wavefront slope matrix;

[0060] The center coordinates of the sub-aperture image obtained from the above calculation This is further converted into a wavefront slope matrix. Specifically, first determine the center coordinates of the sub-aperture image according to the following formula. Convert to the slope value of the corresponding wavefront spatial sampling point:

[0061] ;

[0062] ;

[0063] in is the focal length of the microlens array, and f is the focal length of the primary lens of the wavefront sensor. The diameter of the pupil. The wavelength of light. express Local coordinates of wavefront sampling points in the direction The wavefront slope, express Local coordinates of wavefront sampling points in the direction The wavefront slope.

[0064] Then and Concatenate them sequentially into a single long vector:

[0065] ,

[0066] in, Sub-aperture image after masking The total number of effective pixels within. express Local coordinates of wavefront sampling points in the direction The wavefront slope, express Local coordinates of wavefront sampling points in the direction The wavefront slope.

[0067] S107 concatenates the wavefront slope matrices of all calibration images into a response matrix with the highest spatial sampling rate. Repeat the above process for the calibration image to obtain... slope vectors , Indicates the first The wavefront slope matrices are concatenated to form the response matrix with the highest spatial sampling rate. For the regional calibration method, For the number of deformable mirror actuators, in the mode calibration method, Input Zernike pattern number.

[0068] In one embodiment, see Figure 3 As shown, the slope consistency evaluation within the downsampling window includes:

[0069] S201 divides the sub-aperture coordinate matrix into non-overlapping downsampling windows;

[0070] Receive target space sampling rate input by user Calculate the downsampling window size , This represents the number of pixels corresponding to each sub-aperture. The sub-aperture coordinate matrix corresponding to the light intensity distribution matrix is ​​divided into... Non-overlapping downsampling windows , The set of image pixel indices corresponding to the sub-aperture coordinate matrix is ​​denoted as .

[0071] S202 calculates the local slope variation coefficients in the x and y directions within each downsampling window; each downsampling window Include indivual Directional slope and indivual Directional slope , The coefficients of variation of the local slope in the x and y directions within each downsampling window are defined as the variance of the slope values ​​within that window, to quantify the severity of local tilt changes in the wavefront.

[0072] ,

[0073] ,

[0074] in and These represent the x and y directions within the downsampling window, respectively. The arithmetic mean of the slopes. Let be the local slope variation coefficient in the x-direction. y is the local slope variation coefficient in the y direction.

[0075] S203 compares the local slope variation coefficients in the x and y directions with the variation threshold. If there is a local slope variation coefficient greater than the variation threshold, it is determined that the wavefront target space sampling rate is not suitable for the current wavefront distribution.

[0076] In one embodiment, downsampling the response matrix at the highest spatial sampling rate to generate a slope vector for the target spatial sampling rate includes:

[0077] Based on the uniformity of light intensity distribution, the average value of the slope in the x-direction and the average value of the slope in the y-direction of the target space sampling rate for each calibration image are calculated by selecting either the uniform averaging mode or the light intensity weighted averaging mode, and then concatenated into a long vector. The corresponding long vectors of all calibration images are then concatenated to form the target space sampling rate response matrix.

[0078] The target space sampling rate of each calibrated image is... The position parameter is defined as:

[0079] Uniform averaging mode (suitable for scenarios with uniform light intensity distribution or high signal-to-noise ratio):

[0080] ,

[0081] This represents the average slope of the target space sampling rate in the x-direction of the calibrated image. This represents the average slope of the target space sampling rate of the calibrated image in the y-direction.

[0082] Light intensity weighted average mode (suitable for scenarios with strong turbulence or uneven light intensity distribution, improving noise immunity in low light intensity areas):

[0083] = ;

[0084] ;

[0085] In light intensity weighted mode, light intensity below the system noise threshold is automatically shielded. pixels: if In the weighted calculation, its weight is reset to zero and removed from the denominator; system noise threshold. Determined through dark field noise statistics;

[0086] right The above operations are performed on each calibration image to obtain the average value of the x-direction slope of the target space sampling rate and the average value of the y-direction slope of the target space sampling rate for each calibration image. Each calibration image is then sequentially concatenated into a long vector. :

[0087] ;

[0088] in, Indicates the first Zhang's calibration image at the sampling rate of The mean of the wavefront slope in the direction, Indicates the first Zhang's calibration image at the sampling rate of The mean of the wavefront slope in the direction;

[0089] Will Tensor Vector By sequentially splicing the data, a target space sampling rate response matrix is ​​formed. .

[0090] It should be noted that in this disclosure, relational terms such as "first" and "second" are used merely to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0091] The above description is merely a specific embodiment of this disclosure, enabling those skilled in the art to understand or implement it. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of this disclosure. Therefore, this disclosure is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features claimed herein.

Claims

1. A method for calibrating the multi-space sampling rate response matrix of a deformable mirror, characterized in that, The method includes: The response matrix with the highest spatial sampling rate is obtained by using a checkerboard algorithm on the voltage calibration image acquired by the optical wavefront sensor. Receive the wavefront target spatial sampling rate input by the user, and calculate the downsampling window size of the response matrix with the highest spatial sampling rate; The response matrix of the highest spatial sampling rate is downsampled to generate the slope vector of the target spatial sampling rate, and the slope consistency within the downsampling window is evaluated. Based on the evaluation results, it is determined whether the target spatial sampling rate of the wavefront is suitable for the current wavefront distribution. If it is not suitable, the target spatial sampling rate of the wavefront is increased. If the conditions are met, the slope vectors of all wavefront target space sampling rates are concatenated to generate the target space sampling rate response matrix.

2. The method for calibrating the multi-space sampling rate response matrix of a deformable mirror according to claim 1, characterized in that, The response matrix with the highest spatial sampling rate is obtained using the checkerboard algorithm, including: Each calibration image is segmented based on the sub-aperture of the microlens array to obtain a sub-aperture image; Within the central rectangular region of each sub-aperture image, an inscribed circle binary mask is generated to obtain the sub-aperture image after applying the mask; Find the maximum grayscale value in the set of pixel values ​​corresponding to the same local coordinates in each sub-aperture image; Record the maximum grayscale value and the center coordinates of the sub-aperture image where the maximum grayscale value is located; By traversing all local coordinates, the corresponding maximum gray values ​​are used to form a light intensity distribution matrix, and the center coordinates of the sub-aperture image where the maximum gray value is located are used to form a sub-aperture coordinate matrix. Convert the sub-aperture coordinate matrix into a wavefront slope matrix; The wavefront slope matrices of all the calibration images are concatenated to form the response matrix with the highest spatial sampling rate.

3. The method for calibrating the multi-space sampling rate response matrix of a deformable mirror according to claim 2, characterized in that, The slope consistency assessment within the downsampling window includes: Divide the sub-aperture coordinate matrix into non-overlapping downsampling windows; Calculate the local slope variation coefficients in the x and y directions within each downsampling window; The local slope variation coefficients in the x and y directions are compared with the variation threshold. If there is a local slope variation coefficient greater than the variation threshold, it is determined that the wavefront target space sampling rate is not suitable for the current wavefront distribution.

4. The method for calibrating the multi-space sampling rate response matrix of a deformable mirror according to claim 3, characterized in that, The calculation of the local slope variation coefficients in the x and y directions within each downsampling window includes: calculating the variance of the slope values ​​in the x and y directions within the downsampling window as the local slope variation coefficients in the x and y directions within the downsampling window, respectively.

5. The method for calibrating the multi-space sampling rate response matrix of a deformable mirror according to claim 3, characterized in that, The slope vector for the target spatial sampling rate is generated by downsampling the response matrix at the highest spatial sampling rate, including: Based on the uniformity of light intensity distribution, the average value of the slope in the x-direction and the average value of the slope in the y-direction of the target space sampling rate for each calibration image are calculated by selecting either the uniform averaging mode or the light intensity weighted averaging mode, and then concatenated into a long vector. The corresponding long vectors of all calibration images are then concatenated to form the target space sampling rate response matrix.

6. The method for calibrating the multi-space sampling rate response matrix of a deformable mirror according to claim 1, characterized in that, The calibration images are obtained by using a region calibration method or a mode calibration method. A unit voltage or a Zernike mode voltage is applied to each deformable mirror actuator individually, and the calibration images of the optical wavefront sensor for the voltage are recorded one by one.

Images