A machine vision based method for aligning optical coupling devices

Abstract

The application relates to a machine vision-based light coupling device alignment method, which comprises the following steps: extracting real-time parameter values of refractive index variation, polarization fluctuation and phase delay according to optical data collected by a system; inputting the extracted real-time parameter values of the refractive index variation, the polarization fluctuation and the phase delay into a pre-established multidimensional parameter model for joint modeling; if the refractive index variation exceeds a preset threshold, adjusting the focal length of the system, otherwise keeping the current focal length; calculating a cumulative phase delay effect based on the recalculated parameter values, judging whether the phase delay affects imaging clarity, and if so, performing phase correction; reevaluating system performance indicators according to optical data after phase correction to ensure system stability; optimizing the multidimensional parameter model based on the reevaluated system performance indicators, and obtaining an optimal solution of optical parameters through multiple iteration calculations.

Description

Technical Field

[0001] This application relates to the field of laser measurement instrument technology, and specifically to a machine vision-based optical coupling device alignment method. Background Technology

[0002] In machine vision and optical coupling systems, the multi-dimensional aggregation optimization of optical parameters is a complex technical problem. During the alignment process, the system needs to consider the dynamic changes of multiple optical parameters such as refractive index, polarization state and phase delay. These parameters affect each other during zooming, making it difficult to achieve a balance between aberrations and light energy loss.

[0003] Specifically, changes in refractive index affect the propagation path of light, while fluctuations in polarization state alter the energy distribution of light, and differences in phase delay lead to a decrease in image quality. Joint modeling of these parameters requires finding the optimal solution in a multidimensional space, but the coupling of parameters makes the optimization process exceptionally complex.

[0004] During zooming, as the system focal length increases, the gradient change in refractive index exacerbates aberrations, while abnormal fluctuations in polarization state further amplify light energy loss. Simultaneously, the cumulative effect of phase delay leads to reduced image sharpness. Furthermore, convergence control algorithms under multi-objective constraints need to simultaneously satisfy multiple performance indicators. Achieving the optimal solution in a multi-dimensional parameter space is a core challenge currently facing this technology. Summary of the Invention

[0005] In order to solve the problems existing in the prior art, the purpose of this application is to provide a machine vision-based optical coupling device alignment method.

[0006] The machine vision-based optical coupling device alignment method described in this application includes the following steps:

[0007] Step S101: Extract real-time parameter values ​​of refractive index change, polarization fluctuation and phase delay based on the optical data collected by the system;

[0008] Step S102: Input the extracted real-time parameter values ​​of refractive index change, polarization fluctuation and phase delay into the multi-dimensional parameter model for joint modeling;

[0009] Step S103: Calculate the interaction between refractive index change, polarization fluctuation and phase delay using a multi-dimensional parameter model, and determine whether the refractive index change exceeds a preset threshold based on the calculation results;

[0010] Step S104: If the change in refractive index exceeds a preset threshold, adjust the system focal length; otherwise, maintain the current focal length.

[0011] Step S105: Recalculate the real-time parameter values ​​of refractive index change, polarization fluctuation and phase delay based on the optical data after focal length adjustment;

[0012] Step S106: For the recalculated parameter values, an aberration compensation algorithm is used to perform aberration compensation in order to reduce light energy loss;

[0013] Step S107: Calculate the cumulative phase delay effect based on the recalculated parameter values, determine whether the phase delay affects the image sharpness, and if so, perform phase correction.

[0014] Step S108: Re-evaluate the system performance indicators based on the phase-corrected optical data to ensure system stability;

[0015] Step S109: Optimize the multidimensional parameter model based on the re-evaluated system performance indicators, and obtain the optimal solution of the optical parameters through multiple iterative calculations.

[0016] Preferably, in step S101, the step of extracting real-time parameter values ​​of refractive index change, polarization fluctuation and phase delay based on the optical data collected by the system includes: acquiring a set of real-time optical data collected by optical sensors; performing feature extraction on the real-time optical data to obtain feature parameters characterizing refractive index change, polarization fluctuation and phase delay; and outputting the feature parameters as real-time parameter values ​​to a multi-dimensional parameter model.

[0017] Preferably, in step S102, the step of inputting the extracted real-time parameter values ​​of refractive index change, polarization fluctuation, and phase delay into a pre-established multidimensional parameter model for joint modeling includes: obtaining historical parameter values ​​of refractive index change, polarization fluctuation, and phase delay; inputting the real-time parameter values ​​and historical parameter values ​​into the multidimensional parameter model; and achieving joint modeling through the mapping relationship in the model to obtain a complete multidimensional optical parameter model.

[0018] Preferably, in step S103, the calculation of the mutual influence relationship between refractive index change, polarization fluctuation and phase delay through the multidimensional parameter model includes: inputting the real-time parameter values ​​of refractive index change, polarization fluctuation and phase delay into the multidimensional optical parameter model; calculating the functional relationship between the three based on the preset mathematical equations in the model; and determining the degree of mutual influence between refractive index change, polarization fluctuation and phase delay according to the functional relationship.

[0019] Preferably, in step S105, the aberration compensation using a preset aberration compensation algorithm for the recalculated parameter values ​​includes: obtaining the recalculated refractive index change, polarization fluctuation, and phase retardation parameter values; determining the degree of deviation between the principal ray and the ideal focal plane based on the above parameter values; selecting corresponding compensation coefficients in the aberration compensation algorithm for different degrees of deviation; and using the compensation coefficients to correct the optical imaging data to obtain the compensated optical data.

[0020] Preferably, in step S107, the calculation of the cumulative phase delay effect based on the recalculated parameter values ​​includes: acquiring a set of continuously acquired refractive index change, polarization fluctuation and phase delay parameter values; calculating the phase delay difference between two adjacent sets of parameter values; accumulating multiple sets of phase delay differences to obtain the cumulative phase delay amount; and determining whether the cumulative phase delay amount affects the imaging sharpness.

[0021] Preferably, in step S108, the re-evaluation of system performance indicators based on phase-corrected optical data includes: acquiring a set of phase-corrected test optical data; extracting statistical features of the test optical data, including mean, variance, and signal-to-noise ratio; comparing the statistical features with preset performance indicator thresholds; and determining whether the optical system meets the set stability requirements based on the comparison results.

[0022] Preferably, in step S109, the optimization of the multidimensional parameter model based on the re-evaluated system performance indicators includes: acquiring system performance indicator data from multiple evaluations; dynamically adjusting the weights of each parameter in the multidimensional parameter model according to the performance indicator data; searching for the optimal combination of model parameters through an iterative optimization algorithm; and generating a simplified and efficient multidimensional optical parameter model using the optimal combination of model parameters.

[0023] The machine vision-based optical coupling device alignment method described in this application has the advantage of extracting real-time parameter values ​​of refractive index change, polarization fluctuation, and phase delay from acquired optical data and inputting them into a multi-dimensional parameter model for joint modeling. Based on the model calculation results, the invention determines whether the refractive index change exceeds a preset threshold and adjusts the system focal length accordingly. Subsequently, the invention performs aberration compensation and phase correction on the adjusted optical data to reduce light energy loss and ensure image clarity. Finally, based on the re-evaluated system performance indicators, the invention optimizes the multi-dimensional parameter model and obtains the optimal solution for the optical parameters through multiple iterative calculations. This method enables real-time optimization of optical system parameters, effectively improving system performance and stability, and providing important support for high-precision optical applications. Attached Figure Description

[0024] Figure 1This application describes a machine vision-based optical coupling device alignment method. Figure 1 ;

[0025] Figure 2 This application describes a machine vision-based optical coupling device alignment method. Figure 2 . Detailed Implementation

[0026] like Figures 1-2 As shown, the machine vision-based optical coupling device alignment method described in this application includes the following steps:

[0027] Step S101: Extract real-time parameter values ​​of refractive index change, polarization fluctuation and phase delay based on the optical data collected by the system;

[0028] Step S102: Input the extracted real-time parameter values ​​of refractive index change, polarization fluctuation and phase delay into the multi-dimensional parameter model for joint modeling;

[0029] Step S103: Calculate the interaction between refractive index change, polarization fluctuation and phase delay using a multi-dimensional parameter model, and determine whether the refractive index change exceeds a preset threshold based on the calculation results;

[0030] Step S104: If the change in refractive index exceeds a preset threshold, adjust the system focal length; otherwise, maintain the current focal length.

[0031] Step S105: Recalculate the real-time parameter values ​​of refractive index change, polarization fluctuation and phase delay based on the optical data after focal length adjustment;

[0032] Step S106: For the recalculated parameter values, an aberration compensation algorithm is used to perform aberration compensation in order to reduce light energy loss;

[0033] Step S107: Calculate the cumulative phase delay effect based on the recalculated parameter values, determine whether the phase delay affects the image sharpness, and if so, perform phase correction.

[0034] Step S108: Re-evaluate the system performance indicators based on the phase-corrected optical data to ensure system stability;

[0035] Step S109: Optimize the multidimensional parameter model based on the re-evaluated system performance indicators, and obtain the optimal solution of the optical parameters through multiple iterative calculations.

[0036] like Figures 1-2 As shown, in step S101, real-time parameter values ​​of refractive index change, polarization fluctuation and phase delay are extracted based on the optical data collected by the system.

[0037] Further, in step S101, optical data is acquired to obtain the original optical signal recorded by the system;

[0038] Analyze optical signals to separate basic information such as light intensity, wavelength, and phase;

[0039] Calculate the initial refractive index change based on the light intensity and wavelength data;

[0040] Extract polarization state information and analyze the polarization fluctuation parameters of light waves;

[0041] The phase delay value of the light wave is determined by using phase information;

[0042] If the system focal length is adjusted, the adjusted optical data will be reacquired.

[0043] Based on the adjusted optical data, recalculate the refractive index change.

[0044] Based on the adjusted optical data, update the polarization fluctuation and phase delay parameters;

[0045] The obtained parameters of refractive index change, polarization fluctuation and phase delay are input into a multi-dimensional parameter model for joint modeling.

[0046] Specifically, in step S101, optical data is acquired to obtain the original optical signal recorded by the system, including acquiring a light signal with a wavelength of 532nm through a spectrometer with a sampling frequency of 100Hz;

[0047] The optical signal is analyzed to separate the basic information of light intensity, wavelength and phase. The Fourier transform algorithm is used to decompose the signal into intensity spectrum and phase spectrum.

[0048] Based on the light intensity and wavelength data, the initial refractive index change is calculated. The refractive index is then calculated using Snell's law in conjunction with the light intensity change, including a refractive index change of 0.0015 when the light intensity changes by 10%.

[0049] Polarization state information is extracted, and the polarization fluctuation parameters of the light wave are analyzed. The degree of polarization is calculated to be 0.85 and the polarization angle is 45 degrees using the Stokes parameters.

[0050] The phase delay value of the light wave is determined by the phase information, and the phase unwrapping algorithm is used to obtain the phase delay as π / 2.

[0051] If the system focal length is adjusted, the adjusted optical data will be reacquired, including data reacquisition after the focal length is adjusted from 50mm to 75mm.

[0052] Based on the adjusted optical data, the refractive index change was recalculated, and the refractive index change was found to be 0.0018 using the corrected light intensity and wavelength data.

[0053] Based on the adjusted optical data, the polarization fluctuation and phase delay parameters are updated, and the degree of polarization is recalculated to be 0.88 and the phase delay to be π / 3.

[0054] The obtained parameters of refractive index change, polarization fluctuation and phase delay are input into a multidimensional parameter model for joint modeling. A multivariate linear regression algorithm is used to establish a model of the relationship between refractive index, polarization and phase.

[0055] like Figures 1-2 As shown, in step S102, the extracted real-time parameter values ​​of refractive index change, polarization fluctuation and phase delay are input into the multi-dimensional parameter model for joint modeling.

[0056] Further, in step S102, optical data is acquired, and the original parameter values ​​of refractive index change, polarization fluctuation and phase delay are extracted;

[0057] Based on the optical data after focal length adjustment, the real-time parameter values ​​of refractive index change, polarization fluctuation and phase delay are recalculated;

[0058] The extracted real-time parameter values ​​of refractive index change, polarization fluctuation and phase delay are preprocessed to eliminate noise and outliers;

[0059] The preprocessed real-time parameter values ​​of refractive index change, polarization fluctuation and phase delay are input into the input layer of the multidimensional parameter model;

[0060] The hidden layer of a multidimensional parameter model is used to extract features from the input parameter values ​​and jointly model them to obtain a multidimensional feature vector.

[0061] Based on the multidimensional feature vectors, calculate the weights and biases in the model to determine the correlation between the parameters;

[0062] Through the model output layer, the joint modeling results of refractive index change, polarization fluctuation and phase delay are obtained;

[0063] If the joint modeling results meet the preset threshold, a state assessment report of the optical system will be generated.

[0064] Based on the optical system's condition assessment report, update the parameter weights and biases in the multidimensional parameter model.

[0065] Specifically, in step S102, when acquiring optical data, a spectrometer is used to obtain the light intensity distribution at a sampling frequency of 1000 times per second, and the original parameter values ​​of refractive index change, polarization fluctuation and phase delay are extracted by Fourier transform, wherein the refractive index change ranges from 1.33 to 1.45, the polarization fluctuation amplitude is from 0.01 to 0.05, and the phase delay ranges from 0 to 2π.

[0066] Based on the optical data after focal length adjustment, the real-time parameter values ​​of refractive index change, polarization fluctuation, and phase delay are recalculated using a Gaussian fitting algorithm to ensure data accuracy reaches 10%. -4 ;

[0067] The extracted real-time parameter values ​​are preprocessed by using wavelet transform to eliminate noise and Z-score normalization to remove outliers, so that the data distribution has a mean of 0 and a variance of 1.

[0068] The preprocessed real-time parameter values ​​of refractive index change, polarization fluctuation and phase delay are input into the input layer of the multidimensional parameter model. The number of neurons in the input layer is 128, and the activation function is ReLU.

[0069] The hidden layer of the multidimensional parameter model is used to extract features and jointly model the input parameter values. The hidden layer contains 3 layers, with 256, 128 and 64 neurons in each layer, respectively. The weights are optimized using the backpropagation algorithm to obtain the multidimensional feature vector.

[0070] Based on the multidimensional feature vector, the weights and biases in the model are calculated, and the correlation between the parameters is determined using the gradient descent method. The learning rate is set to 0.001.

[0071] The model output layer uses the Softmax function to obtain the joint modeling results of refractive index change, polarization fluctuation and phase delay. The number of neurons in the output layer is 3.

[0072] If the joint modeling results meet the preset thresholds, i.e., the refractive index change error is less than 0.001, the polarization fluctuation error is less than 0.005, and the phase delay error is less than 0.01, then a state assessment report of the optical system is generated, which includes parameter values ​​and their confidence intervals.

[0073] Based on the optical system's condition assessment report, the Adam optimizer was used to update the parameter weights and biases in the multi-dimensional parameter model, ensuring that the model accuracy was improved to 99.9%.

[0074] like Figures 1-2 As shown, in step S103, the interaction between refractive index change, polarization fluctuation and phase delay is calculated by a multidimensional parameter model, and the refractive index change is judged based on the calculation results to determine whether it exceeds a preset threshold.

[0075] Further, in step S103, the initial parameters of the optical system are obtained, including the raw data of refractive index, polarization fluctuation and phase delay;

[0076] A multi-parameter model is used to model and calculate the interaction between refractive index variation, polarization fluctuation and phase delay;

[0077] Based on the calculation results of the multidimensional parameter model, extract the real-time value of the refractive index change;

[0078] The extracted refractive index change value is compared with a preset threshold to determine whether it exceeds the threshold range.

[0079] If the refractive index change exceeds the preset threshold, adjust the focal length parameter of the optical system and reacquire the optical data.

[0080] Based on the adjusted focal length parameters, the real-time values ​​of refractive index change, polarization fluctuation, and phase delay are recalculated.

[0081] The recalculated values ​​of refractive index change, polarization fluctuation, and phase delay are input into the multidimensional parameter model for joint modeling.

[0082] By combining modeling results, the interaction between refractive index variation, polarization fluctuation and phase delay is updated;

[0083] Based on the updated mutual influence relationships, the optimized parameters of the optical system are determined, and the model calculations are completed.

[0084] Specifically, in step S103, the initial parameters of the optical system include a refractive index of 1.52, polarization fluctuation of 0.03 rad, and phase delay of 0.15 rad, and these raw data are acquired through the data acquisition module;

[0085] Using a multi-parameter model, initial values ​​of refractive index, polarization fluctuation, and phase delay are input, and the interaction between the three is calculated using matrix operations and the least squares method, yielding a refractive index variation coefficient of 0.02.

[0086] Based on the calculation results of the multidimensional parameter model, the real-time value of the refractive index change is extracted as 0.021;

[0087] The extracted refractive index change value of 0.021 is compared with the preset threshold of 0.02, and it is determined that it exceeds the threshold range.

[0088] If the refractive index change exceeds a preset threshold, the focal length parameter of the optical system is adjusted from 10mm to 10.5mm using an algorithm, and the optical data is reacquired.

[0089] Based on the adjusted focal length parameters, the real-time values ​​of refractive index change, polarization fluctuation and phase delay were recalculated using the finite element analysis method, and were 0.019, 0.028 rad and 0.14 rad, respectively.

[0090] The recalculated values ​​are input into the multidimensional parameter model, and Gaussian elimination and iterative optimization algorithms are used for joint modeling.

[0091] By combining the modeling results, the interaction between refractive index variation, polarization fluctuation and phase delay is updated to obtain a new interaction coefficient matrix;

[0092] Based on the updated mutual influence relationship, the optimal parameters of the optical system are determined using the gradient descent method, the model calculation is completed, and the final result is output.

[0093] like Figures 1-2 As shown, in step S104, if the refractive index change exceeds a preset threshold, the system focal length is adjusted; otherwise, the current focal length is maintained.

[0094] Furthermore, in step S104, the real-time parameter values ​​of the refractive index change, polarization fluctuation, and phase delay of the current optical system are obtained through a multi-dimensional parameter model;

[0095] Based on the calculation results of the multidimensional parameter model, the mutual influence relationship between refractive index change, polarization fluctuation and phase delay is determined;

[0096] The system compares a preset threshold with the real-time parameter value of the refractive index change to determine whether the refractive index change exceeds the preset threshold.

[0097] If the change in refractive index exceeds a preset threshold, the system's focus adjustment mechanism is triggered, generating a focus adjustment command.

[0098] If the change in refractive index does not exceed the preset threshold, the current system focal length remains unchanged, and the current state of the optical system is maintained.

[0099] According to the focal length adjustment command, the system focal length adjustment operation is executed to recalibrate the focal length parameters of the optical system;

[0100] The real-time parameter values ​​of refractive index change, polarization fluctuation and phase delay after focal length adjustment are recalculated using a multi-dimensional parameter model.

[0101] The optical data in the multidimensional parameter model is updated using recalculated real-time parameter values ​​to ensure the real-time performance of the model parameters.

[0102] Based on the updated multi-parameter model, we continue to monitor the dynamic changes in refractive index, polarization fluctuations, and phase delay.

[0103] Specifically, in step S104, the real-time parameter values ​​of the refractive index change, polarization fluctuation and phase delay of the current optical system are obtained through a multi-dimensional parameter model, including the calculation of the refractive index change as 1.52, the polarization fluctuation as 0.03 and the phase delay as 0.12 using the finite element analysis method.

[0104] Based on the calculation results of the multidimensional parameter model, the mutual influence relationship between refractive index change, polarization fluctuation and phase delay was determined. The correlation coefficient between refractive index change and polarization fluctuation was 0.85 when analyzed by linear regression model.

[0105] The system compares a preset threshold with the real-time parameter value of the refractive index change to determine whether the refractive index change exceeds the preset threshold. For example, if the preset threshold is 1.50, and the real-time refractive index change is 1.52, then the threshold is exceeded.

[0106] If the change in refractive index exceeds the preset threshold, the system's focal length adjustment mechanism is triggered, generating a focal length adjustment command, including adjusting the focal length from 50mm to 52mm.

[0107] If the change in refractive index does not exceed the preset threshold, the current system focal length remains unchanged, and the current state of the optical system is maintained.

[0108] According to the focal length adjustment command, the system focal length adjustment operation is executed to recalibrate the focal length parameters of the optical system, including adjusting the focal length to 52mm through a precision stepper motor;

[0109] The real-time parameter values ​​of refractive index change, polarization fluctuation and phase delay after focal length adjustment were recalculated using a multi-dimensional parameter model, including a corrected refractive index change of 1.50, polarization fluctuation of 0.02 and phase delay of 0.10.

[0110] The optical data in the multidimensional parameter model is updated using recalculated real-time parameter values ​​to ensure the real-time performance of the model parameters, including updating the refractive index change to 1.50.

[0111] Based on the updated multidimensional parameter model, we continue to monitor the dynamic changes in refractive index, polarization fluctuations, and phase delay, including setting the monitoring frequency to 10 times per second to acquire optical parameter values ​​in real time.

[0112] like Figures 1-2 As shown, in step S105, the real-time parameter values ​​of refractive index change, polarization fluctuation and phase delay are recalculated based on the optical data after focal length adjustment.

[0113] Furthermore, in step S105, based on the optical data after focal length adjustment, the basic parameters of the optical system, such as wavelength, incident angle, and beam intensity, are extracted.

[0114] Using the optical transmission equation and combining fundamental parameters, the propagation path and energy distribution of the light beam in the optical system are calculated.

[0115] By using the refractive index calculation formula and combining the beam propagation path and energy distribution, the real-time parameter values ​​of refractive index change can be obtained.

[0116] Based on the polarization state analysis method, the polarization state change information of the beam during propagation is extracted; using the polarization wave model and combining the polarization state change information, the real-time parameter values ​​of the polarization wave are calculated.

[0117] By using the phase delay calculation formula and combining the beam propagation path and polarization state change information, the real-time parameter value of the phase delay can be obtained.

[0118] Based on the input requirements of the multidimensional parameter model, real-time parameter values ​​of refractive index change, polarization fluctuation and phase delay are integrated;

[0119] A joint modeling method using multidimensional parameter models is employed to perform multidimensional correlation analysis on the integrated real-time parameter values;

[0120] The comprehensive performance index of the optical system at different focal lengths is determined by the output results of the multi-dimensional parameter model.

[0121] Specifically, in step S105, based on the optical data after the focal length is adjusted to 200 mm, basic parameters such as wavelength λ = 532 nm, incident angle θ = 30°, and beam intensity I = 100 mW are extracted;

[0122] Using the optical transmission equation and these parameters, the propagation path of the beam in the optical system is calculated, yielding the energy distribution E(x,y)=I·exp(-(x 2 +y 2 ) / w 2 ), where w is the beam radius;

[0123] By using the refractive index calculation formula n=sinθ / sinθ', combined with the propagation path and energy distribution, the real-time parameter value of refractive index change Δn=0.01 is obtained;

[0124] Based on the polarization state analysis method, the polarization state change information of the beam during propagation is extracted, and the Stokes parameter S = [1, 0.8, 0.6, 0] is obtained;

[0125] Using a polarization wave model and Stokes parameters, the real-time parameter value of polarization wave ΔP = 0.02 was calculated.

[0126] By using the phase delay calculation formula Δφ=(2π / λ)·Δn·d, combined with propagation path and polarization state change information, the real-time parameter value of phase delay Δφ=0.05 radians is obtained;

[0127] Based on the input requirements of the multi-dimensional parameter model, the real-time parameter values ​​of refractive index change Δn, polarization fluctuation ΔP, and phase delay Δφ are integrated to form the input matrix M = [Δn, ΔP, Δφ].

[0128] A joint modeling method using multidimensional parameter models is employed to perform multidimensional correlation analysis on the input matrix, yielding a correlation coefficient matrix R = [[1,0.9,0.8], [0.9,1,0.85], [0.8,0.85,1]].

[0129] Based on the output of the multi-dimensional parameter model, the overall performance index of the optical system at a focal length of 200 mm is determined to be Q = 0.95.

[0130] like Figures 1-2 As shown, in step S106, an aberration compensation algorithm is used to compensate for aberrations in order to reduce light energy loss for the recalculated parameter values.

[0131] Further, in step S106, real-time parameter values ​​of refractive index change, polarization fluctuation and phase delay are obtained based on the optical data after focal length adjustment;

[0132] An aberration compensation algorithm is used to process the recalculated values ​​of refractive index change, polarization fluctuation, and phase delay parameters;

[0133] The distribution of light energy loss in the optical system is calculated using an aberration compensation algorithm.

[0134] Based on the distribution of light energy loss, determine the compensation parameters that need to be adjusted in the aberration compensation algorithm;

[0135] Aberrations in the optical system are compensated using compensation parameters to obtain compensated optical data;

[0136] Based on the compensated optical data, the cumulative phase delay effect is recalculated;

[0137] Determine whether the cumulative phase delay effect exceeds a preset threshold; if it does, perform phase correction.

[0138] By using a phase correction algorithm, the phase delay parameter in the optical system is adjusted to obtain the corrected optical data;

[0139] The imaging sharpness information of the optical system is updated based on the corrected optical data.

[0140] Specifically, in step S106, based on the optical data after focal length adjustment, the refractive index change is obtained through numerical simulation as 1.45 to 1.55, the polarization fluctuation range is ±0.02, and the phase delay is 0.1 to 0.3 radians.

[0141] The Zernike polynomial aberration compensation algorithm is used to fit the recalculated values ​​of refractive index change, polarization fluctuation and phase delay parameters to obtain the aberration distribution matrix;

[0142] The optical energy loss distribution in the optical system is calculated using an aberration compensation algorithm and the finite element method, with the loss value ranging from 0.5% to 2.5%.

[0143] Based on the distribution of light energy loss, determine the compensation parameters that need to be adjusted in the aberration compensation algorithm, and set the compensation coefficients to 0.8 to 1.2;

[0144] Aberrations in the optical system are compensated using compensation parameters. The compensated optical data is obtained through iterative optimization, and the root mean square error of aberrations is reduced to below 0.05.

[0145] Based on the compensated optical data, the cumulative phase delay effect was recalculated using Fourier transform, with a phase delay value of 0.15 to 0.25 radians;

[0146] Determine whether the cumulative phase delay effect exceeds the preset threshold of 0.2 radians. If it does, use the least squares method for phase correction.

[0147] By using a phase correction algorithm, the phase delay parameter in the optical system is adjusted, and the corrected phase delay value is controlled between 0.1 and 0.15 radians.

[0148] Based on the corrected optical data, the imaging sharpness information of the optical system is updated using the point spread function, improving the resolution to 200 line pairs per millimeter.

[0149] like Figures 1-2 As shown, in step S107, the cumulative phase delay effect is calculated based on the recalculated parameter values, and it is determined whether the phase delay affects the image sharpness. If it does, phase correction is performed.

[0150] Further, in step S107, the initial parameter values ​​of the optical system are obtained, including refractive index, polarization state and phase information;

[0151] Based on a multi-dimensional parameter model, the influence of refractive index variation on polarization fluctuations is calculated.

[0152] Based on the calculation results of polarization fluctuations, the cumulative amount of phase delay effect is determined;

[0153] The cumulative amount of phase retardation effect is used to determine whether it affects image sharpness;

[0154] If the phase delay effect affects the image sharpness, a phase correction algorithm is used to correct it.

[0155] Based on the phase-corrected data, the refractive index change of the system is recalculated;

[0156] By recalculating the change in refractive index, it is determined whether it exceeds a preset threshold.

[0157] The system performance indicators are re-evaluated based on the refractive index change that does not exceed the preset threshold.

[0158] The final stability of the system is determined based on the re-evaluated system performance metrics.

[0159] Specifically, in step S107, the initial parameter values ​​of the optical system are obtained, including refractive index 1.45, polarization state 0.8, and phase information π / 3;

[0160] Based on a multi-parameter model, the influence of refractive index change on polarization fluctuation was calculated using the finite element analysis method, and the amplitude of polarization fluctuation was found to be 0.12.

[0161] Based on the calculation results of polarization fluctuations, the cumulative amount of phase delay effect is determined to be 0.05π using an integral algorithm;

[0162] The cumulative amount of phase delay effect is used to determine whether it affects image sharpness using a sharpness evaluation function, with a sharpness threshold of 0.9.

[0163] If the phase delay effect affects the image sharpness, the least squares method is used for phase correction, and the corrected phase value is π / 4.

[0164] Based on the phase-corrected data, the refractive index change of the system was recalculated, and the refractive index change value was found to be 0.02.

[0165] The recalculated change in refractive index is compared with a preset threshold of 0.03 to determine whether it exceeds the preset threshold.

[0166] Based on the refractive index change not exceeding the preset threshold, the system performance indicators, including resolution and contrast, were re-evaluated, and the resolution was found to be 200 lp / mm and the contrast ratio was 95%.

[0167] Based on the reassessed system performance metrics, the final stability of the system was determined to be 98%.

[0168] like Figures 1-2 As shown, in step S108, the system performance indicators are re-evaluated based on the phase-corrected optical data to ensure system stability.

[0169] Further, in step S108, the phase-corrected optical data is acquired, and the phase-corrected optical signal is extracted;

[0170] Calculate the real-time parameter value of the system phase delay based on the phase-corrected optical signal;

[0171] The phase stability of the optical system is analyzed using the real-time parameter value of the system phase delay.

[0172] Based on the phase-corrected optical data, extract the real-time parameter values ​​of refractive index change;

[0173] The range of refractive index fluctuation in the optical system is calculated using real-time parameter values ​​of refractive index changes.

[0174] Based on the range of refractive index fluctuations, determine the refractive index stability of the optical system under different environments;

[0175] Real-time parameter values ​​of polarization fluctuations are extracted using phase-corrected optical data;

[0176] The polarization stability index of the optical system is calculated based on the real-time parameter values ​​of polarization fluctuations.

[0177] The overall performance of an optical system is comprehensively evaluated using refractive index stability, phase stability, and polarization stability indicators.

[0178] Specifically, in step S108, phase-corrected optical data is obtained from the optical system, and the phase-corrected optical signal is extracted using a Fourier transform algorithm, including calculating the phase component with a signal frequency of 500Hz.

[0179] Based on the extracted phase correction signal, the real-time parameter values ​​of the system phase delay are calculated using the least squares method, including a delay time of 0.05ms.

[0180] Based on the phase delay parameter value, the phase stability of the optical system is analyzed by calculating the phase standard deviation as 0.01 rad.

[0181] Real-time parameter values ​​of refractive index variation were extracted from the phase-corrected optical data using the refractive index calculation formula, including refractive index variation ranges from 1.45 to 1.47.

[0182] By using the refractive index variation parameter value and calculating the refractive index fluctuation amplitude as 0.02, the refractive index fluctuation range of the optical system is obtained.

[0183] Based on the range of refractive index fluctuation, the refractive index stability of the optical system under different environments is determined, including a refractive index fluctuation of less than 0.01 when the temperature changes by ±5℃.

[0184] Real-time parameter values ​​of polarization fluctuations, including a polarization degree of 0.95, are extracted from the phase-corrected optical data using Stokes parameters.

[0185] Based on the polarization fluctuation parameter values, calculate the polarization stability index of the optical system, including a polarization angle deviation of less than 0.5°;

[0186] Combining refractive index stability, phase stability, and polarization stability indices, a weighted comprehensive evaluation method is used to calculate the overall performance score of the optical system, with a comprehensive score of 95 points.

[0187] like Figures 1-2 As shown, in step S109, the multi-dimensional parameter model is optimized based on the re-evaluated system performance indicators, and the optimal solution of the optical parameters is obtained through multiple iterative calculations.

[0188] Further, in step S109, the phase-corrected optical data is acquired, and the system performance indicators are extracted;

[0189] Based on the extracted system performance indicators, construct the initial parameter set for the multidimensional parameter model;

[0190] An iterative calculation method is used to adjust the optical parameters in the multidimensional parameter model;

[0191] The interaction between refractive index variation, polarization fluctuation and phase delay was calculated using a multi-dimensional parameter model.

[0192] Determine whether the calculated change in refractive index exceeds a preset threshold.

[0193] If the refractive index change does not exceed the preset threshold, update the optical parameters in the multidimensional parameter model based on the calculation results.

[0194] The repeated iterative calculation process optimizes the updated multidimensional parameter model multiple times.

[0195] The optimal solution for the optical parameters is obtained through multiple iterative calculations;

[0196] Re-evaluate system performance metrics based on the optimal solution to confirm system stability.

[0197] Specifically, in step S109, system performance indicators are extracted from the phase-corrected optical data, including extracting the root mean square value of wavefront error (RMS) as 0.05λ and the Strehl ratio as 0.92.

[0198] Based on the extracted system performance indicators, an initial parameter set for the multidimensional parameter model is constructed, including an initial refractive index of 1.45, a polarization fluctuation range of ±0.1°, and an initial phase delay of π / 4.

[0199] An iterative calculation method, including gradient descent, is used to adjust the optical parameters in the multidimensional parameter model. The learning rate is set to 0.01 and the iteration step size is 100.

[0200] The interaction between refractive index variation, polarization fluctuation and phase delay is calculated using a multi-dimensional parameter model, including solving the covariance matrix of refractive index variation and polarization fluctuation using matrix operations.

[0201] Determine whether the calculated change in refractive index exceeds a preset threshold, which is ±0.01.

[0202] If the refractive index change does not exceed the preset threshold, the optical parameters in the multidimensional parameter model are updated according to the calculation results, including adjusting the refractive index to 1.452 and correcting the polarization fluctuation range to ±0.08°.

[0203] The iterative calculation process involves repeatedly optimizing the updated multidimensional parameter model, recording the parameter convergence status in each iteration.

[0204] Through multiple iterative calculations, the optimal solution for the optical parameters was obtained, including a final refractive index of 1.453, a polarization fluctuation range of ±0.07°, and a phase delay of π / 3.

[0205] The system performance metrics were reassessed based on the optimal solution, including calculating that the root mean square value of the wavefront error decreased to 0.03λ and the Strelby ratio increased to 0.95, thus confirming the system stability.

[0206] For those skilled in the art, various other corresponding changes and modifications can be made based on the technical solutions and concepts described above, and all such changes and modifications should fall within the protection scope of the claims of this application.

Claims

1. A machine vision-based method for aligning optically coupled devices, characterized in that, Includes the following steps: Step S101: Extract real-time parameter values ​​of refractive index change, polarization fluctuation and phase delay based on the optical data collected by the system; Step S102: Input the real-time parameter values ​​of refractive index change, polarization fluctuation and phase delay extracted in step S101 into the multi-dimensional parameter model for joint modeling. Step S103: Calculate the interaction between refractive index change, polarization fluctuation and phase delay using the multidimensional parameter model in step S102, and determine whether the refractive index change exceeds a preset threshold based on the calculation results. Step S104: If the change in refractive index exceeds a preset threshold, adjust the system focal length; otherwise, maintain the current focal length. Step S105: Recalculate the real-time parameter values ​​of refractive index change, polarization fluctuation and phase delay based on the optical data after focal length adjustment in step S104. Step S106: For the parameter values ​​recalculated in step S105, an aberration compensation algorithm is used to perform aberration compensation in order to reduce light energy loss. Step S107: Calculate the cumulative phase delay effect based on the parameter values ​​recalculated in step S106, determine whether the phase delay affects the image sharpness, and if so, perform phase correction. Step S108: Re-evaluate the system performance indicators based on the phase-corrected optical data from step S107. Step S109: Optimize the multidimensional parameter model based on the system performance indicators re-evaluated in step S108, and obtain the optimal solution of the optical parameters through iterative calculation.

2. The machine vision-based optical coupling device alignment method according to claim 1, characterized in that, The step S101 of extracting real-time parameter values ​​includes: acquiring real-time optical data collected by optical sensors, separating light intensity, wavelength and phase information through Fourier transform, calculating the refractive index change value by combining Snell's law, analyzing polarization fluctuations using Stokes parameters, and determining the phase delay value through a phase unpacking algorithm.

3. The machine vision-based optical coupling device alignment method according to claim 1, characterized in that, The joint modeling in step S102 specifically includes: Real-time and historical parameter values ​​are input into a multidimensional parameter model. A functional relationship between refractive index, polarization, and phase is established through a multiple linear regression algorithm, and the model weights are optimized based on the gradient descent method.

4. The machine vision-based optical coupling device alignment method according to claim 1, characterized in that, The step S103 of adjusting the system focal length includes: When the refractive index change exceeds the threshold, a focal length adjustment command is generated and the focal length is adjusted by a stepper motor. The adjustment range is ±5% of the current focal length value, and optical data is re-acquired after adjustment for parameter verification.

5. The machine vision-based optical coupling device alignment method according to claim 1, characterized in that, The aberration compensation algorithm in step S104 is the Zernike polynomial algorithm, which includes: The aberration distribution matrix is ​​fitted based on the refractive index change, polarization fluctuation and phase delay parameter values. The optical energy loss distribution is calculated by finite element analysis, and the compensated optical data is iteratively optimized based on the compensation coefficient.

6. The machine vision-based optical coupling device alignment method according to claim 1, characterized in that, The phase correction in step S105 includes: The cumulative phase delay effect is corrected by the least squares method, and the phase delay value is controlled within ±10% of the preset threshold. The image sharpness is updated by the point spread function.

7. The machine vision-based optical coupling device alignment method according to claim 1, characterized in that, The optimization of the multidimensional parameter model in step S106 includes: Based on the system performance indicators after phase correction, the Adam optimizer is used to dynamically adjust the model parameter weights, and the gradient descent method is used for iterative optimization until the parameter convergence error is less than 0.5%.

8. The machine vision-based optical coupling device alignment method according to claim 2, characterized in that, The sampling frequency of the Fourier transform is from 100Hz to 1000Hz, and the calculation accuracy of the Stokes parameters is ±0.

01.

9. The machine vision-based optical coupling device alignment method according to claim 3, characterized in that, The hidden layer of the multidimensional parameter model contains three neural network layers with 256, 128, and 64 neurons respectively, and the activation function is ReLU. The output layer uses the Softmax function to generate the joint modeling result.

10. The machine vision-based optical coupling device alignment method according to claim 5, characterized in that, The compensation coefficients of the Zernike polynomial algorithm range from 0.8 to 1.2, and the root mean square error of the aberration after compensation is less than 0.05.

Images