Optical phased array phase calibration method and apparatus, electronic device, and storage medium

Abstract

The application relates to an optical phased array phase calibration method and device, an electronic device and a storage medium, and is applied to the field of optical detection. The optical phased array phase calibration method comprises the following steps: inputting a to-be-calibrated phase image into an optical phased array system, and obtaining a random disturbance phase image corresponding to the to-be-calibrated phase image; the random disturbance phase image is obtained by performing phase disturbance within a preset range on the to-be-calibrated phase image. The to-be-calibrated phase image and the random disturbance phase image are input into a target convolutional neural network model to determine a target phase error of the to-be-calibrated phase; the target convolutional neural network model is obtained by training a preset convolutional neural network model based on a far-field training image pair; and the initial emission electric field of the to-be-calibrated phase image is calibrated based on the target phase error. The application solves the problem that the accuracy of the existing optical phased array phase error calibration needs to be improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of optical detection, and in particular to an optical phased array phase calibration method and device, an electronic device and a storage medium. BACKGROUND

[0002] An optical phased array (OPA) is a technology that uses multiple optical transmitting units or optical receiving units to control the direction of a light beam based on the principle of phased array, and has the advantages of fast scanning speed, high robustness and low sensitivity to environmental vibration. However, in practical applications, in order to improve the resolution of the OPA, a large array structure with more array elements is usually used to reduce the beam width. However, the OPA using a large array of array elements, especially an array of thermal optical modulators, is prone to introduce phase error problems due to unevenness of device technology.

[0003] The existing related calibration algorithm generally compensates for the phase error through a random parallel gradient descent algorithm, but does not consider the problem of non-unique solution of the phase error in the optical phase space, resulting in the need for further improvement in the accuracy of phase error calibration of the optical phased array.

[0004] At present, there is no effective solution to the problem that the accuracy of the existing phase error calibration of the optical phased array needs to be improved in the related art. SUMMARY

[0005] An optical phased array phase calibration method, device, electronic device and storage medium are provided in the present embodiment to solve the problem that the accuracy of the existing phase error calibration of the optical phased array needs to be improved in the related art.

[0006] In a first aspect, an optical phased array phase calibration method is provided in the present embodiment, applied to an optical phased array system, the optical phased array system comprising an optical phased array and a target convolutional neural network model; the method comprising:

[0007] inputting a phase image to be calibrated into the optical phased array system to obtain a random perturbation phase image corresponding to the phase image to be calibrated; the phase image to be calibrated is generated according to an initial transmitting electric field in the optical phased array; the random perturbation phase image is obtained by perturbing the phase of the phase image to be calibrated within a preset range;

[0008] inputting the to-be-calibrated phase image and the random perturbation phase image into a target convolutional neural network model to determine a target phase error of the to-be-calibrated phase; the target convolutional neural network model is obtained by training a preset convolutional neural network model based on a far-field training image pair; the far-field training image pair includes a far-field training image generated according to an optical phased array and a far-field random perturbation phase image;

[0009] calibrating an initial emission electric field of the to-be-calibrated phase image based on the target phase error.

[0010] In some embodiments, the inputting the to-be-calibrated phase image and the random perturbation phase image into a target convolutional neural network model to determine a target phase error of the to-be-calibrated phase includes:

[0011] performing grayscale processing on the to-be-calibrated phase image and the random perturbation phase image respectively to obtain a to-be-calibrated phase grayscale image and a random perturbation phase grayscale image;

[0012] based on a preset sampling strategy, inputting the to-be-calibrated phase grayscale image and the random perturbation phase grayscale image after sampling segmentation into the target convolutional neural network model;

[0013] predicting the target phase error of the to-be-calibrated phase image through a nonlinear mapping relationship between a far-field image and a phase in the target convolutional neural network model and based on the to-be-calibrated phase grayscale image and the random perturbation phase grayscale image after sampling segmentation; the nonlinear mapping relationship between the far-field image and the phase is trained according to the far-field training image pair.

[0014] In some embodiments, the optical phased array system further includes a light splitting component, a lens component, and an optical phased array chip; and the method further includes:

[0015] when a laser light source generates an output light beam after passing through the optical phased array, dividing the output light beam into a first light beam and a second light beam through the light splitting component;

[0016] focusing the first light beam through the lens component to obtain the far-field training image corresponding to the first light beam;

[0017] determining an image electric signal corresponding to the second light beam through the optical phased array chip; and based on a preset perturbation strategy, randomly perturbing a phase of the image electric signal corresponding to the second light beam to obtain an image perturbation electric signal corresponding to the second light beam;

[0018] determining a second light beam far-field training image and a second light beam far-field random perturbation phase image corresponding to the image electric signal and the image perturbation electric signal respectively.

[0019] In some embodiments, the preset convolutional neural network model comprises a residual module and a sampling module; the method further comprises:

[0020] inputting the far-field training image pair with the phase mapping range limitation and the random phase distribution corresponding to the far-field training image pair into the preset convolutional neural network;

[0021] extracting features of the far-field training image pair through the residual module to obtain training image features;

[0022] down-sampling the training image features through the sampling module to output target training image features; the size of the target training image features is set to half of the training image features;

[0023] determining a predicted phase error according to the target training image features;

[0024] determining a target convolutional neural network model according to the predicted phase error and the real compensation phase corresponding to the far-field training image pair based on a preset normalization evaluation function.

[0025] In some embodiments, the determining a target convolutional neural network model according to the predicted phase error and the real compensation phase corresponding to the far-field training image pair based on a preset normalization evaluation function comprises:

[0026] calculating a phase difference value between the predicted phase error and the real compensation phase;

[0027] when it is judged that the phase difference value is a multiple of a first phase, inputting the far-field training image pair and the random phase distribution corresponding to the far-field training image pair into the preset convolutional neural network model for training until the phase difference value is a multiple of a second phase;

[0028] when it is judged that the phase difference value is a multiple of a second phase, determining a training termination time in the preset convolutional neural network model, and determining the target convolutional neural network model after the current time reaches the training termination time; the second phase is greater than the first phase.

[0029] In some embodiments, the determining a target convolutional neural network model comprises:

[0030] training the preset convolutional neural network model according to a preset mean square error loss function to obtain a target convolutional neural network model.

[0031] In some embodiments, the calibrating the initial emission electric field of the to-be-calibrated phase image based on the phase error further comprises:

[0032] determine an initial transmit electric field generating the phase image to be calibrated;

[0033] calibrate the initial transmit electric field according to the phase error to obtain a target transmit electric field;

[0034] determine a far-field intensity of the target transmit electric field after phase calibration according to the phase error based on a preset far-field intensity calculation strategy;

[0035] generate a target phase image according to the target transmit electric field when it is judged that the far-field intensity of the target transmit electric field meets a preset intensity threshold; the target phase image includes an image generated after the phase image to be calibrated is calibrated.

[0036] In a second aspect, an optical phased array phase calibration system is provided in the embodiment, and the system includes an input module, an error determination module and a calibration module.

[0037] The input module is configured to input a phase image to be calibrated into an optical phased array system to obtain a random disturbance phase image corresponding to the phase image to be calibrated; the phase image to be calibrated is generated according to an initial transmit electric field in the optical phased array; and the random disturbance phase image is obtained by performing phase disturbance within a preset range on the phase image to be calibrated.

[0038] The error determination module is configured to input the phase image to be calibrated and the random disturbance phase image into a target convolutional neural network model to determine a target phase error of the phase image to be calibrated; the target convolutional neural network model is trained based on a preset convolutional neural network model and a pair of far-field training images; and the pair of far-field training images includes a far-field training image and a far-field random disturbance phase image generated according to the optical phased array.

[0039] The calibration module is configured to calibrate an initial transmit electric field of the phase image to be calibrated based on the target phase error.

[0040] In a third aspect, an electronic device is provided in the embodiment, which includes a memory, a processor and a computer program stored in the memory and executable on the processor; and the processor implements the optical phased array phase calibration method of the first aspect when executing the computer program.

[0041] In a fourth aspect, a storage medium is provided in the embodiment, which stores a computer program executable by a processor to implement the optical phased array phase calibration method of the first aspect.

[0042] Compared with the related art, the optical phased array phase calibration method, device, electronic device and storage medium provided in the embodiment can train a convolutional neural network through a far-field diffraction image and an image of random phase disturbance, obtain a nonlinear mapping relationship of the far-field diffraction image and the phase, and make the target convolutional neural network model obtained through training avoid the symmetric phase in the optical phased array, thereby avoiding the problem of multiple solutions of the phase. Meanwhile, the target convolutional neural network model obtained through loss iteration training can effectively improve the accuracy of determining the phase error that needs to be compensated, thereby improving the accuracy of phase error calibration of the optical phased array.

[0043] The details of one or more embodiments of the present application are presented in the following drawings and description to make other features, objects and advantages of the present application more clear and simple. BRIEF DESCRIPTION OF DRAWINGS

[0044] The drawings described herein are used to provide further understanding of the present application, constitute a part of the present application, and the illustrative embodiments of the present application and the description thereof are used to explain the present application, and do not constitute an improper limitation on the present application. In the drawings:

[0045] Figure 1 is a hardware structure block diagram of a terminal of the optical phased array phase calibration method provided by the embodiment of the present application;

[0046] Figure 2 is a flowchart of the optical phased array phase calibration method provided by the embodiment of the present application;

[0047] Figure 3 is a flowchart of determining a far-field training image pair in the embodiment of the present application;

[0048] Figure 4 is a structural schematic diagram of the optical phased array system provided by the embodiment of the present application;

[0049] Figure 5 is a structural block diagram of the optical phased array phase calibration system of the embodiment. DETAILED DESCRIPTION

[0050] In order to more clearly understand the purpose, technical scheme and advantages of the present application, the present application is described and explained in combination with the drawings and embodiments.

[0051] Unless otherwise defined, technical terms or scientific terms used in the present application shall have the same meaning as those commonly understood by a person of ordinary skill in the art to which the present application belongs. The terms "one", "a", "an", "the", "these", and similar terms in the present application do not mean "only one" or "exactly one", but can mean "one or more". The terms "include", "contain", "have", and any variant thereof in the present application are intended to cover the non-exclusive inclusion; for example, a process, method, system, product or device containing a series of steps or modules (units) is not limited to the listed steps or modules (units), but can include steps or modules (units) not listed, or can include other steps or modules (units) inherent to the process, method, product or device. The terms "connect", "connected", "couple" and the like in the present application are not limited to physical or mechanical connections, but can include electrical connections, whether direct or indirect. The term "multiple" in the present application means two or more. The term "and / or" describes the association relationship between the associated objects, which means that there can be three relationships, for example, "A and / or B" can mean that A exists alone, A and B exist together, and B exists alone. Generally, the character " / " means that the objects associated before and after are in an "or" relationship. The terms "first", "second", "third" and the like in the present application are only used to distinguish similar objects, and do not represent a specific order of the objects.

[0052] The method embodiments provided in the present embodiment can be executed in a terminal, a computer or a similar computing device. For example, the method embodiments are executed on a terminal, Figure 1 is a hardware structure block diagram of a terminal of the optical phased array phase calibration method provided by the embodiments of the present application. As shown in Figure 1 , the terminal can include one or more (only one is shown in Figure 1 ) processor 102 and memory 104 for storing data, wherein the processor 102 can include but not limited to processing devices such as microprocessor MCU or programmable logic device FPGA. The above terminal can also include transmission equipment 106 for communication function and input / output equipment 108. Those skilled in the art can understand that Figure 1 The structure shown is only schematic, which does not limit the structure of the above terminal. For example, the terminal can include more or less components than Figure 1 shown, or have a different configuration from Figure 1 shown.

[0053] The memory 104 can be used to store computer programs, such as software programs of application software and modules, such as a computer program corresponding to the optical phased array phase calibration method in the embodiment. The processor 102 can execute various functional applications and data processing, i.e., implement the method described above, by running the computer program stored in the memory 104. The memory 104 can include a high-speed random access memory, and can further include a non-volatile memory, such as one or more magnetic storage devices, flash memories, or other non-volatile solid-state memories. In some examples, the memory 104 can further include memories remotely arranged with respect to the processor 102, which can be connected to the terminal through a network. Examples of the network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and a combination thereof.

[0054] The transmission device 106 is configured to receive or send data via a network. The network includes a wireless network provided by a communication provider of the terminal. In an example, the transmission device 106 includes a network interface controller (NIC) that can be connected to other network devices through a base station to communicate with the Internet. In an example, the transmission device 106 can be a radio frequency (RF) module configured to communicate with the Internet in a wireless manner.

[0055] The core of the optical phased array (OPA) technology is to synthesize a directional light beam or directionally receive an incident light beam by controlling the phases of a series of closely arranged optical transmitting or receiving elements. However, for an optical phased array with a large number of elements, processing deviation will inevitably occur in the processing process, resulting in phase errors in the output light field of the transmitting antenna elements. At present, in order to calibrate the phase errors and achieve high-precision deflection of the OPA output light beam, many important algorithms have been proposed to calibrate the output light field, such as genetic algorithm (GA), hill climbing (HC), and stochastic parallel gradient descent (SPGD) algorithm.

[0056] Among them, the hill climbing algorithm is a local search algorithm used to find a local optimal solution. It starts from an initial solution and then tries to find a better solution through a series of improvement steps. If a new solution is better than the current one, the algorithm will move to this new solution and continue searching. The process is repeated until no better solution is found. However, the hill climbing algorithm is difficult to calibrate for large-scale OPAs because it scans the channel signal one by one, which is too time-consuming. Genetic algorithms can be used to calibrate large-scale operational amplifiers, but their convergence time is too long (feedback period, phase compensation time is longer), and the efficiency of phase calibration needs to be improved. The stochastic parallel gradient descent algorithm is suitable for continuous wave and high repetition rate pulse combination, but the number of steps required for convergence (feedback period) is about ten times the number of combined beams. If there is one sampling point per pulse and one step or less per sampling point, this process becomes too slow for pulses with kilohertz repetition rates and combinations of hundreds of beams. The stochastic parallel gradient descent algorithm is hindered by only obtaining information from one detector, which cannot indicate which beam or direction to adjust, and it also introduces noise through the jitter and search process. Moreover, the traditional stochastic parallel gradient descent algorithm is prone to local optimal solutions, resulting in slow convergence.

[0057] At the same time, for the optical phased array array, the array arrangement is regular, and the sub-beam at the center position is taken as the origin, and the initial phase value of 180° symmetry in space can be obtained. The phase values of the sub-beams on the left and right sides of the center sub-beam And The same phase difference will be generated, resulting in the same far-field image. In the image recognition of CNN, the far-field image obtained has a non-unique phase value. As the input of the model, the far-field image cannot accurately reflect the phase information of the OPA, and the CNN model cannot predict a phase value to match the same optimal solution, and it also cannot guarantee that the training model learns the mapping relationship from the far-field image to the optical phase. For different symmetrical arrays, different array phases will produce the same far-field image. In practice, in order to distinguish different phase values, such as phase solution 1 and phase solution 2, an additional light splitting device (wavefront modulator and beam splitter) is usually used, but this will increase the difficulty of system integration, especially for large array optical phased arrays, which is not the optimal solution. Since the optical phase space of the optical phased array is center-symmetric, there may be non-unique solutions when calibrating the phase, and the current phase calibration algorithm is not sufficient to meet the requirements of OPA beam phase calibration.

[0058] Therefore, in the embodiment, an optical phased array phase calibration method is provided, applied to an optical phased array system including an optical phased array and a target convolutional neural network model; by adopting a non-uniform array structure to distinguish two array phase unique solutions, the problem of phase ambiguity is solved. On the one hand, in the process of optimizing the array structure, the array structure is trained and optimized to realize the problem of low sidelobe of the optical phased array; on the other hand, according to the non-uniform array structure between the training image including the original image and the random perturbation phase image and the far field image, the existence of conjugate symmetric phase can be avoided, and the problem of phase ambiguity can be avoided. Figure 2 is a flowchart of the optical phased array phase calibration method provided by the embodiment of the present application, as shown in Figure 2 , the flow includes the following steps:

[0059] In step S210, input the phase image to be calibrated into the optical phased array system, and obtain a random perturbation phase image corresponding to the phase image to be calibrated; the phase image to be calibrated is generated according to the initial emission electric field in the optical phased array; the random perturbation phase image is obtained by perturbing the phase in a preset range on the phase image to be calibrated.

[0060] In the process of realizing far field imaging by the laser light source via the optical phased array, if beam deviation or misalignment, or abnormal beam shape occurs, at this time, it indicates that there is a phase error in the phase image to be calibrated, therefore, the specific array phase solution, i.e. the phase error, needs to be determined, and the array phase of the initial emission electric field is calibrated according to the phase error, and then the generated phase image to be calibrated is calibrated.

[0061] When the phase image to be calibrated needs to be calibrated, the phase image to be calibrated needs to be input into the optical phased array system; the optical phased array system includes a target convolutional neural network model trained according to the far field training image and the far field random perturbation phase image, containing the nonlinear mapping relationship between the far field image and the phase.

[0062] After the phase image to be calibrated is input into the optical phased array system, the phase of the initial emission electric field generating the phase image to be calibrated is obtained, and the random perturbation phase image generated by the initial emission electric field after phase perturbation in a preset range is obtained. Here, the phase perturbation in the preset range can be determined according to the historical phase error, or can be determined according to the array phase of the actual optical phased array and laser light source, which is not limited here.

[0063] The optical phased array system here can include an optical phased array, a phase control module for phase control of the optical phased array, and a lens assembly for subsequent imaging, etc., which is not limited here.

[0064] Step S220, inputting the to-be-calibrated phase image and the random perturbation phase image into a target convolutional neural network model to determine a target phase error of the to-be-calibrated phase; the target convolutional neural network model is obtained by training a preset convolutional neural network model based on a far-field training image pair; the far-field training image pair includes a far-field training image generated according to the optical phased array and a far-field random perturbation phase image.

[0065] In this step, after obtaining the to-be-calibrated image and the random perturbation phase image, the to-be-calibrated phase image and the random perturbation phase image are input into the target convolutional neural network model. According to the target convolutional neural network model, the phase error of the to-be-calibrated phase image is predicted to obtain the target phase error.

[0066] The convolutional neural network (CNN) is a kind of feedforward neural network containing convolution calculation and having a deep structure; the general architecture of the CNN usually includes an input layer, a convolution layer, an activation function, a pooling layer, a fully connected layer, an output layer, and a normalization layer in some architectures to accelerate the training process and stabilize the gradient. The preset convolutional neural network model is determined based on the above convolutional neural network.

[0067] Step S230, calibrating the initial emission electric field of the to-be-calibrated phase image based on the target phase error.

[0068] In this step, after inputting the to-be-calibrated phase image into the optical phased array system, the target phase error of the to-be-calibrated phase image is determined according to the trained target convolutional neural network model; the array phase of the initial emission electric field forming the to-be-calibrated phase image is calibrated according to the target phase error to obtain the calibrated target emission electric field.

[0069] Through the above steps, the preset convolutional neural network model is trained by using the far-field training image pair including the far-field training image generated according to the optical phased array and the far-field random perturbation phase image, i.e., the far-field diffraction image and the far-field training image with random phase perturbation, to obtain the target convolutional neural network model including the nonlinear mapping relationship between the far-field diffraction image and the phase, so that the target convolutional neural network model obtained by training can avoid the symmetric phase in the optical phased array and further avoid the problem of multiple solutions of the phase; at the same time, the target convolutional neural network model obtained by loss iteration training can effectively improve the accuracy of determining the phase error that needs to be compensated, and further improve the accuracy of phase error calibration of the optical phased array.

[0070] In some embodiments, the calibrating the initial transmit electric field of the to-be-calibrated phase image based on the target phase error further includes: determining the initial transmit electric field of the to-be-calibrated phase image; calibrating the initial transmit electric field according to the target phase error to obtain a target transmit electric field; determining the far-field intensity of the target transmit electric field after phase calibration according to the target phase error based on a preset far-field intensity calculation strategy; and generating a target phase image according to the target transmit electric field when it is determined that the far-field intensity of the target transmit electric field meets a preset intensity threshold. The target phase image includes an image generated after the to-be-calibrated image is calibrated.

[0071] The preset far-field intensity calculation strategy includes determining based on an element factor of the target transmit electric field, a phase of an optical path in the element, and the target phase error that needs to be compensated, which is not specifically limited here. When the calculated far-field intensity of the target transmit electric field meets the preset intensity threshold, it means that there is no phase error in the target transmit electric field, and a more concentrated far-field spot can be generated by the optical phased array.

[0072] In some embodiments, the method for obtaining the far-field training image pair of the target convolutional neural network model trained can be determined based on a preset perturbation strategy in an optical phased array system. The optical phased array system includes an optical phased array and a target convolutional neural network model, and further includes a light splitting component, a lens component, and an optical phased array chip. The chip of the optical phased array is controlled by a phase control module to realize far-field imaging. The phase control module is not shown in the figure and can be an FPGA (Field-Programmable Gate Array), a DSP (digital signal processor), an MCU (Microcontroller Unit), etc., which is not specifically limited here. The optical phased array chip is specifically a silicon-based optical phased array chip.

[0073] When the laser light source passes through the optical phased array controlled by the phase control module, an output light beam is generated, and the far-field training image pair is determined according to the output light beam.

[0074] Specifically, Figure 3 is a flowchart for determining the far-field training image pair in the embodiments of the present application, and reference is made to Figure 3 The method for determining the far-field training image pair includes the following steps S310 to S340.

[0075] In step S310, an output light beam is generated when the laser light source passes through the optical phased array, and the output light beam is divided into a first light beam and a second light beam via a light splitting component.

[0076] Step S320, focusing the first light beam through the lens assembly to obtain a far-field training image corresponding to the first light beam.

[0077] Wherein, the first light beam is focused through the lens assembly to obtain the far-field training image, the performance index of the target convolutional neural network model is determined according to the far-field training image, and the preset convolutional neural network model is continuously corrected and trained to obtain the target convolutional neural network meeting the performance index.

[0078] Step S330, determining an image electric signal corresponding to the second light beam through the optical phased array chip; and randomly disturbing the phase of the image electric signal corresponding to the second light beam based on a preset disturbance strategy to obtain an image disturbance electric signal corresponding to the second light beam.

[0079] Wherein, the preset disturbance strategy can be adding Gaussian noise to the image electric signal corresponding to the second light beam, and the Gaussian noise is determined based on the phase error existing in the actual imaging process.

[0080] Step S340, respectively determining a second light beam far-field training image and a second light beam far-field random disturbance phase image corresponding to the image electric signal and the image disturbance electric signal.

[0081] By dividing the output light beam generated by the laser light source through the optical phased array into two light beams, focusing and disturbance training are performed on the two light beams respectively, so that the target convolutional neural network finally trained is more robust, and the phase error of the light beam element in the optical phased array is improved to obtain an output stable emission electric field.

[0082] Further, the method for determining the to-be-calibrated phase image and the random disturbance phase image is consistent with the method for determining the far-field training image pair, which will not be described again.

[0083] In some embodiments, the to-be-calibrated phase image and the random disturbance phase image are input into the target convolutional neural network model to determine the target phase error of the to-be-calibrated phase, including: respectively performing gray processing on the to-be-calibrated phase image and the random disturbance phase image to obtain a to-be-calibrated phase gray image and a random disturbance phase gray image; based on a preset sampling strategy, the to-be-calibrated phase gray image and the random disturbance phase gray image are input into the target convolutional neural network model after being segmented by sampling; the target phase error of the to-be-calibrated phase image is predicted through a nonlinear mapping relationship between a far-field image and a phase in the target convolutional neural network model and based on the to-be-calibrated phase gray image and the random disturbance phase gray image segmented by sampling; and the nonlinear mapping relationship between the far-field image and the phase is obtained by training according to the far-field training image pair.

[0084] By performing gray processing on the image, the data amount and the computational complexity can be greatly reduced. By performing sampling segmentation on the image, the image resolution can be reduced, and the data amount to be processed can be reduced, so as to accelerate the processing speed and reduce the demand for computational resources. By performing gray processing on the to-be-calibrated phase image and the randomly disturbed phase image and the sampling segmentation method, more features in the extracted image are extracted, which is beneficial to further improve the accuracy of determining the nonlinear mapping relationship between the far-field image and the phase, and further improve the accuracy and efficiency of subsequent prediction of the target phase error.

[0085] In the process of training the preset convolutional neural network model according to the far-field training image pair, the far-field training image pair is subjected to gray processing and sampling segmentation, and then the nonlinear mapping relationship between the far-field image and the phase is trained.

[0086] In some embodiments, the preset convolutional neural network model includes a residual module and a sampling module; the optical phased array phase calibration method further includes: inputting the far-field training image pair subjected to phase mapping range limitation and the random phase distribution corresponding to the far-field training image pair into the preset convolutional neural network; extracting features of the far-field training image pair by the residual module to obtain training image features; performing down-sampling on the training image features by the sampling module to output target training image features; the size of the target training image features is set to half of the size of the training image features; determining a predicted phase error according to the target training image features; determining a target convolutional neural network model based on the preset normalization evaluation function and according to the predicted phase error and the real compensation phase corresponding to the far-field training image pair.

[0087] In some embodiments, the preset convolutional neural network model includes a residual module and a sampling module; the optical phased array phase calibration method further includes: inputting the far-field training image pair subjected to phase mapping range limitation and the random phase distribution corresponding to the far-field training image pair into the preset convolutional neural network; extracting features of the far-field training image pair by the residual module to obtain training image features; performing down-sampling on the training image features by the sampling module to output target training image features; the size of the target training image features is set to half of the size of the training image features; determining a predicted phase error according to the target training image features; determining a target convolutional neural network model based on the preset normalization evaluation function and according to the predicted phase error and the real compensation phase corresponding to the far-field training image pair.

[0088] In some embodiments, the first phase can be an odd multiple of π, and the second phase can be an even multiple of π. For example, the first phase is π, and the second phase is 2π.

[0089] In some embodiments, the target convolutional neural network model is determined by training the preset convolutional neural network model according to a preset mean square error loss function to obtain the target convolutional neural network model.

[0090] The present embodiment is described and explained below through specific embodiments.

[0091] In the present embodiment, a fast convergence algorithm for calibrating the phase error of large array optical phased array beam deflection is proposed. The optical phased array is controlled by a novel iterative method based on machine learning, and the beam phase is calibrated by an experimentally calibrated numerical simulation model. The array of the optical phased array here can be 32, 64, 128, etc., and here a 64 array is taken as an example. Based on the recognition of the far-field interference pattern of the optical phased array beam, a neural network is trained to detect the phase error. There may be non-unique solutions in the optical phase space, and the neural network needs to be trained within a limited phase perturbation, which can reduce the number of samples required for training. When the phase perturbation is small, the trained neural network can make the phase value of the emission electric field in the optical phased array within the range of 2π, and its iterative scheme can be randomly detected and quickly converged within the training range. To solve the problem of non-unique solutions in the prior art, double-frame far-field imaging images are used as input data sets to improve the generalization ability of CNN, and the initial phase compensation result is optimized.

[0092] Convolutional Neural Network (CNN) is a powerful image processing tool, and its diverse architecture brings the potential of complex mapping relationship fitting. The convolution kernel in its network structure is a two-dimensional structure, and the use of shared weights between network parameters can improve the feature learning ability and computational efficiency of the neural network for image data. The CNN is applied to the optical phased array phase calibration system to realize the fast phase locking between array beams, and a large number of far-field images with phase labels can be used to train the CNN, which can accurately establish the nonlinear mapping relationship between the far-field diffraction image and the system phase, and can directly predict the phase error in the current system.

[0093] With the above CNN technology, fast phase retrieval and calibration can be achieved in the OPA system. Compared with the traditional SPGD algorithm commonly used at present, this technology has great advantages: (1) Compared with the SPGD algorithm, which requires more steps to recover the aberration for a lower performance evaluation function, the CNN has better retrieval effect for different initial results, and can obtain higher far-field output in one step and is more robust to different input ranges; (2) The CNN is a wavefront sensor that can directly predict the OPA aberration without additional iterative optimization; (3) When facing an OPA with a large array structure, only the output layer parameters need to be fine-tuned to adapt to the number of sub-beams, and the convergence time can remain unchanged.

[0094] The above steps help to accurately control the behavior of the optical phased array to achieve the desired far-field performance, such as beam focusing or shaping in a specific direction, and other applications. Through continuous iteration and optimization, the algorithm gradually reduces the difference between the actual output and the target, achieving the optimization purpose.

[0095] Further, the residual neural network method is adopted, which has a jump connection structure, i.e. shortcut structure, which can effectively avoid gradient diffusion / explosion and inhibit network degradation. Due to the shortcut connection, even in the case of a very deep network, the gradient can be effectively returned to the previous layer through these direct paths. In this way, the gradient vanishing problem commonly seen in traditional deep networks is avoided, because the gradient no longer needs to be transmitted layer by layer. Network degradation refers to the phenomenon that as the network depth increases, the performance may become worse. ResNet overcomes this challenge by allowing the network to learn residual mapping instead of complete bottom mapping, making the learning process easier. The network constructed using residual blocks can achieve better accuracy than the network without residual blocks under the same conditions.

[0096] The target convolutional neural network model in the embodiment is based on a basic residual neural network module: a conventional residual module extracts features from the input without changing the size (height and width) of the feature map during the calculation process; another residual module used in the embodiment performs down-sampling during feature extraction, and the size of the output feature is half of the input feature. The shortcut connection of the residual network can propagate the features extracted by the shallow layer to the deep layer, enhancing the expression ability of the neural network through feature reuse while avoiding the risk of overfitting.

[0097] In the embodiment, an optimized neural network architecture is used to extract features from the input layer and down-sample them. The shortcut connection of the residual network can extend the features extracted by the shallow layer to the deep layer, thereby enhancing the expression ability of the neural network through feature reuse.

[0098] The proportion factor depends on the demand for the target and the hardware resource constraint, and a smaller proportion factor can be used to retain more detailed information. However, with the increase of the proportion factor, although the model size can be effectively reduced, important information can be lost, thereby affecting the final accuracy; for example, a proportion factor of 0.5 is used in the downsampling module.

[0099] Further, batch normalization is used in the convolutional neural network model to make the training more stable. Swish is selected as the activation function here because it has a smooth nonlinearity and better performance in stabilizing training and improving accuracy. The expression of the Swish activation function is as follows:

[0100] ;

[0101] wherein x represents an input value, and sigma(x) is a Sigmoid function; in order to convert a nonlinear problem into a linear problem, x represents the input value of the activation function, and is the result of the sum of the weighted input received by the current neuron and the bias term.

[0102] Further, the convolutional neural network model is converged by setting a neural network evaluation function. The phase error is compensated to evaluate the OPA phase calibration performance, and then the performance of the CNN under different parameters is compared to determine the training termination time. In view of the characteristics of the array beam aberration, the normalized phase difference cosine is used to evaluate the network performance in the embodiment, and the formula of the neural network evaluation function is as follows:

[0103] ;

[0104] wherein N represents the number of beam arrays of the optical phased array, represents the phase to be compensated, represents the predicted phase in the neural network optimization process.

[0105] When the smaller indicates that the prediction of the neural network for the beam far-field phase is closer to the test actual value. The value of ranges from 0 to 1, and the smaller the value, the closer the prediction value of the network to the true value. When the difference between and is an integer multiple of 2pi, will be 0, and the prediction value is equal to the true value. When the difference between and is an odd multiple of pi, will be 1.

[0106] In one specific embodiment, the method of training the neural network includes: using 20000 sets of OPA far-field imaging pictures with random phase distribution as the input source of the neural network, wherein each set of pictures includes an initial image and a random phase perturbation image. 15000 sets of far-field images and corresponding random phases (phase mapping range [0, 2π]) are used for training, and the remaining 5000 sets are used for testing. Further, the MSE (Mean Squared Error) is used as the loss function to train the network, and the average value of the evaluation function is evaluated with the change of the training steps. For example, the test is performed once every 10 training steps. The neural network evaluation function is as follows The piston phase residual is described. In the actual experiment, after 260 epochs of training, the phase error can be reduced to 0.005. The lower the value, the closer the network output value to the real value, the more accurate the network output, and the more accurate the image generated after phase error compensation. The neural network here is the convolutional neural network model in the foregoing embodiment.

[0107] The optical phased array system includes a field-programmable gate array (FPGA), an analog-to-digital converter (ADC), a digital-to-analog converter (DAC), an electro-optic phase modulator, and an acousto-optic modulator; it also involves pseudo-random bit sequence (PRBS), frequency hopping (FH), and heterodyne frequency technologies. PRBS is a commonly used test signal pattern that provides a standardized method to evaluate system performance, such as bit error rate, during system testing and calibration. Frequency hopping is a radio communication technology that quickly switches operating frequencies to avoid interference and eavesdropping, optimizing the quality and security of communication links. Heterodyne technology involves mixing the received signal with a reference signal generated by a local oscillator to produce a lower frequency signal for subsequent processing. In a heterodyne receiver, this process allows efficient amplification and filtering of high-frequency signals.

[0108] Figure 4 is a structural schematic diagram of the optical phased array system provided by the embodiment of the present application. Referring to Figure 4To verify the ability of CNN to calibrate the phase, a 36-element optical phased array system is constructed to generate the dataset. The packaged 64-element (8x8) OPA chip is controlled by the FPGA phase control module to realize far-field imaging. After the laser light source is input into the FPGA phase control module, the light beam output by the OPA is split into two beams by a beam splitter. One part of the light beam, i.e., the first light beam, passes through a focusing lens with a focal length of 10 cm to the focal plane of the photodetector, which is used to generate CNN performance indicators and detect far-field images for continuous training, and ultimately corrects random dynamic aberrations. The other part of the light beam, i.e., the second light beam, is used to obtain a dataset with a new assigned random phase perturbation image. In this way, 200 groups of datasets are generated. The second light beam generates a corresponding image electrical signal through a charge-coupled device. After that, based on a preset perturbation strategy, the phase of the image electrical signal corresponding to the second light beam is randomly perturbed through a programmable array, such as an FPGA, to obtain a perturbed image electrical signal corresponding to the second light beam. The second light beam far-field training image and the second light beam far-field random perturbation phase image corresponding to the image electrical signal and the perturbed image electrical signal are respectively subjected to grayscale processing and sampling segmentation. As shown in the figure, the image size obtained by sampling segmentation is 32x32x2, 16x16x64, 4x4x256, and 2x2x512, respectively. After that, the sampling segmented second light beam far-field training image and the sampling segmented second light beam far-field random perturbation phase image are respectively input into a preset convolutional neural network model to be trained or a target convolutional neural network model trained to perform model training or model application. P1-P6 represent the outputs of multiple layers in the convolutional neural network model.

[0109] Unlike the far-field image simulated by the first light beam, the dataset including the second light beam far-field training image and the second light beam far-field random perturbation phase image may be affected by phase noise, thereby causing a distribution difference between experimental data and training data, which may affect the performance of the model. Therefore, by sampling the size of the input image during training and adding Gaussian noise to the initial image to simulate noise, the trained model is more robust, and the beam error in the OPA system can be corrected very quickly, and a stable and good light field is ultimately obtained.

[0110] Compared with using the approximate linear relationship between the mean square sum of aberration gradient and the second moment of far-field intensity distribution to process the far-field grayscale image, in this embodiment, the CNN is used to subsample the OPA far-field grayscale image to extract more features. For the deterministic mapping problem, the phase is derived from the far-field coherent diffraction model to quickly find the phase error to be corrected.

[0111] For the identification of far-field diffraction patterns, a convolutional neural network is used to identify the mapping function between the far-field diffraction image and the beam phase error. The mapping function can be achieved by training the neural network with an interference imaging mode with a known phase disturbance. After identifying the mapping function, the neural network can identify the original beam diffraction pattern and find the corresponding phase error of any given pattern in a very short deployment.

[0112] Exemplarily, a neural network architecture is constructed for an 8x8 array beam phase and 16x16 far-field diffraction image mapping, with 256 neurons as an input layer, including 64 predicted beam phases connected to many inputs, which generate an output signal when the input exceeds a set threshold. The neuron is the sum of all input signals multiplied by a weight matrix that describes the mapping function. The weight matrix determines the sensitivity of the neuron to each input. The training of the network is to adjust the sensitivity of each neuron to its input, which refers to the degree of influence (change) of the array random phase on the far-field image, that is, the degree of change of the evaluation function of the neural network. When training the network, the weight of each neuron in the network is adjusted to change the output of the neural network to match the true output (i.e., the known phase disturbance).

[0113] Based on the target convolutional neural network model based on the training, the target phase error of the phase image to be calibrated is determined and calibrated, and the array beam diffraction model of the optical phased array beam control system can be established to determine whether the phase error is eliminated after compensation by the target phase error, and to determine the far-field intensity after compensation.

[0114] Wherein, the array beam diffraction model of the optical phased array beam control system is established, and the emission electric field of the optical phased array can be expressed as:

[0115] ;

[0116] Wherein, , λ represents the wavelength of light, m represents the number of elements in the optical phased array, is the light field amplitude, is the far-field observation distance, and the phase of the element at m is , the directional pattern of a single element is represented by the element factor , represents the incident angle, represents the direction angle.

[0117] Generally, the array is usually assumed to be a uniform array with element spacing d and amplitude A=1, so the emission electric field of the above optical phased array can be simplified as:

[0118] ;

[0119] Wherein, m represents the number of array elements in the optical phased array, For the far-field observation distance, = md, the phase of the mth array element is The directional diagram of a single array element is represented by the array element factor , represents the incident angle, represents the direction angle.

[0120] When there is a phase error, the light field phase needs to be compensated, and then the compensated far-field intensity can be represented as:

[0121] ;

[0122] Wherein: is the phase noise of the mth light path, is the phase that needs to be compensated. represents the phase of the mth light path.

[0123] In the actual system, if there is no phase noise, .

[0124] When the above equation does not hold, it means that there is actually phase noise, and then the phase noise of the light field of the optical phased array needs to be compensated, so that the far-field energy can be more concentrated.

[0125] In this embodiment, an optical phased array phase calibration system is also provided, which is used to implement the above embodiments and preferred embodiments, and has been described above and will not be repeated. The terms "module", "unit", "sub-unit" and the like used below can be a combination of software and / or hardware that can implement the predetermined function. Although the devices described in the following embodiments are preferably implemented in software, hardware, or a combination of software and hardware can also be implemented and conceived.

[0126] Figure 5 The structure block diagram of the optical phased array phase calibration system of this embodiment is shown in Figure 5 , which includes an input module 10, an error determination module 20 and a calibration module 30.

[0127] The input module 10 is used to input the phase image to be calibrated into the optical phased array system, and to obtain a random disturbance phase image corresponding to the phase image to be calibrated; the phase image to be calibrated is generated according to the initial emission electric field in the optical phased array; the random disturbance phase image is obtained by performing phase disturbance within a preset range on the phase image to be calibrated;

[0128] The error determination module 20 is configured to input the to-be-calibrated phase image and the random perturbation phase image into a target convolutional neural network model to determine a target phase error of the to-be-calibrated phase; the target convolutional neural network model is obtained by training a preset convolutional neural network model based on a far-field training image pair; the far-field training image pair includes a far-field training image generated according to the optical phased array and a far-field random perturbation phase image.

[0129] The calibration module 30 is configured to calibrate the initial transmit electric field of the to-be-calibrated phase image based on the target phase error.

[0130] It should be noted that each of the above modules can be a functional module or a program module, and can be implemented by software or hardware. For the modules implemented by hardware, each of the above modules can be located in the same processor; or each of the above modules can also be located in different processors in any combination.

[0131] In this embodiment, an electronic device is also provided, including a memory and a processor, the memory stores a computer program, and the processor is configured to execute the computer program to perform the steps in any of the above method embodiments.

[0132] Optionally, the electronic device can further include a transmission device and an input / output device, wherein the transmission device is connected to the processor, and the input / output device is connected to the processor.

[0133] Optionally, in this embodiment, the processor can be configured to execute the following steps through the computer program:

[0134] S1, input the to-be-calibrated phase image into the optical phased array system to obtain a random perturbation phase image corresponding to the to-be-calibrated phase image; the to-be-calibrated phase image is generated according to an initial transmit electric field in the optical phased array; the random perturbation phase image is obtained by performing phase perturbation within a preset range on the to-be-calibrated phase image.

[0135] S2, input the to-be-calibrated phase image and the random perturbation phase image into a target convolutional neural network model to determine a target phase error of the to-be-calibrated phase; the target convolutional neural network model is obtained by training a preset convolutional neural network model based on a far-field training image pair; the far-field training image pair includes a far-field training image generated according to the optical phased array and a far-field random perturbation phase image.

[0136] S3, calibrate the initial transmit electric field of the to-be-calibrated phase image based on the target phase error.

[0137] It should be noted that the specific examples in the present embodiment can refer to the examples described in the above embodiments and optional implementation manners, which will not be described herein again.

[0138] In addition, in combination with the optical phased array phase calibration method provided in the above embodiments, a storage medium can also be provided in the present embodiment to implement. The storage medium has a computer program stored thereon; the computer program is executed by a processor to implement any one of the optical phased array phase calibration methods in the above embodiments.

[0139] It should be understood that the specific embodiments described herein are intended to explain, not limit, the application. Based on the embodiments provided in the present application, all other embodiments obtained by those of ordinary skill in the art without creative labor are within the scope of protection of the present application.

[0140] Obviously, the drawings are only some examples or embodiments of the present application, and those of ordinary skill in the art can also apply the present application to other similar situations without creative labor. In addition, it can be understood that although the work done in the development process may be complex and long, some design, manufacture or production changes made by those of ordinary skill in the art according to the technical content disclosed in the present application are only routine technical means and should not be regarded as insufficient disclosure of the present application.

[0141] The word "embodiment" in the present application means that the specific features, structures or characteristics described in combination with the embodiments can be included in at least one embodiment of the present application. The appearance of this phrase in various places in the specification does not necessarily mean the same embodiment, nor does it mean independence or alternative to other embodiments. Those of ordinary skill in the art can clearly or implicitly understand that the embodiments described in the present application can be combined with other embodiments without conflict.

[0142] The above-described embodiments only express several implementation manners of the present application, which are described in detail and specifically, but should not be understood as limitations to the patent protection scope. It should be noted that for those of ordinary skill in the art, without departing from the concept of the present application, several modifications and improvements can be made, which are within the scope of protection of the present application. Therefore, the protection scope of the present application should be subject to the appended claims.

Claims

1. A method for calibrating the phase of an optical phased array, characterized in that, The method is applied to an optical phased array system, which includes an optical phased array and a target convolutional neural network model; the method includes: The phase image to be calibrated is input into the optical phased array system to obtain a random perturbation phase image corresponding to the phase image to be calibrated; the phase image to be calibrated is generated based on the initial emission electric field in the optical phased array; the random perturbation phase image is obtained by perturbing the phase image to be calibrated within a preset range; the phase perturbation within the preset range is determined based on historical phase error, or based on the array phase of the actual optical phased array and laser source. The phase image to be calibrated and the randomly perturbed phase image are input into a target convolutional neural network model to determine the target phase error of the phase to be calibrated. The target convolutional neural network model is trained on a preset convolutional neural network model based on far-field training image pairs. The target convolutional neural network model includes a nonlinear mapping relationship between the far-field image and the phase. The far-field training image pairs include a far-field training image generated based on an optical phased array and a far-field randomly perturbed phase image obtained by adding Gaussian noise to the electrical signal of the far-field training image. The initial emission electric field of the phase image to be calibrated is calibrated based on the target phase error to obtain the target emission electric field; when it is determined that the far-field intensity of the target emission electric field meets the preset intensity threshold, a target phase image is generated based on the target emission electric field; the target phase image includes the image generated after calibrating the phase image to be calibrated. The method further includes: determining a predicted phase error based on a preset convolutional neural network model; determining a target convolutional neural network model based on a preset normalized evaluation function and the true compensation phase corresponding to the predicted phase error and the far-field training image pair; wherein, determining the target convolutional neural network model based on the preset normalized evaluation function and the true compensation phase corresponding to the predicted phase error and the far-field training image pair includes: calculating the phase difference between the predicted phase error and the true compensation phase; when it is determined that the phase difference is a multiple of a first phase, inputting the far-field training image pair and the random phase distribution corresponding to the far-field training image pair into the preset convolutional neural network model for training until the phase difference is a multiple of a second phase; when it is determined that the phase difference is a multiple of the second phase, determining the training termination time in the preset convolutional neural network model, and determining the target convolutional neural network model after the current time reaches the training termination time; the first phase is π, and the second phase is 2π.

2. The optical phased array phase calibration method according to claim 1, characterized in that, The step of inputting the phase image to be calibrated and the randomly perturbed phase image into the target convolutional neural network model to determine the target phase error of the phase to be calibrated includes: The phase image to be calibrated and the random perturbation phase image are respectively processed into grayscale to obtain the grayscale image of the phase to be calibrated and the grayscale image of the random perturbation phase; Based on a preset sampling strategy, the phase grayscale image to be calibrated and the randomly perturbed phase grayscale image are sampled and segmented respectively, and then input into the target convolutional neural network model; The nonlinear mapping relationship between the far-field image and the phase in the target convolutional neural network model is used to predict the target phase error of the phase image to be calibrated based on the sampled and segmented grayscale image of the phase to be calibrated and the randomly perturbed grayscale image; the nonlinear mapping relationship between the far-field image and the phase is obtained by training the far-field training image pair.

3. The optical phased array phase calibration method according to claim 1, characterized in that, The optical phased array system further includes a beam splitter assembly, a lens assembly, and an optical phased array chip; the method further includes: When the laser source passes through the optical phased array, it generates an output beam, which is then divided into a first beam and a second beam by a beam splitter. The first beam is focused by a lens assembly to obtain the far-field training image corresponding to the first beam; The image electrical signal corresponding to the second beam is determined by an optical phased array chip; based on a preset perturbation strategy, the phase of the image electrical signal corresponding to the second beam is randomly perturbed to obtain the image perturbation electrical signal corresponding to the second beam. The second beam far-field training image and the second beam far-field random perturbation phase image corresponding to the image electrical signal and the image perturbation electrical signal are determined respectively.

4. The optical phased array phase calibration method according to claim 1, characterized in that, The preset convolutional neural network model includes a residual module and a sampling module; the method further includes: The far-field training image pairs with phase mapping range limitation and the random phase distribution corresponding to the far-field training image pairs are input into a preset convolutional neural network; The residual module is used to extract features from the far-field training image pairs to obtain the training image features; The sampling module downsamples the training image features to output target training image features; the size of the target training image features is set to half that of the training image features. The prediction phase error is determined based on the features of the target training image; Based on a preset normalized evaluation function, the target convolutional neural network model is determined according to the predicted phase error and the true compensation phase corresponding to the far-field training image pair.

5. The optical phased array phase calibration method according to claim 4, characterized in that, Determining the target convolutional neural network model includes: The preset convolutional neural network model is trained according to the preset mean squared error loss function to obtain the target convolutional neural network model.

6. The optical phased array phase calibration method according to claim 1, characterized in that, The method further includes calibrating the initial emission electric field of the phase image to be calibrated based on the target phase error to obtain the target emission electric field, and also includes: Determine the initial emission electric field that generates the phase image to be calibrated; The initial emission electric field is calibrated based on the phase error to obtain the target emission electric field; Based on a preset far-field intensity calculation strategy, the far-field intensity of the target emitted electric field after phase calibration is determined according to the phase error.

7. An optical phased array phase calibration system, characterized in that, The system includes: an input module, an error determination module, and a calibration module; The input module is used to input the phase image to be calibrated into the optical phased array system to obtain a random perturbation phase image corresponding to the phase image to be calibrated; the phase image to be calibrated is generated based on the initial emission electric field in the optical phased array; the random perturbation phase image is obtained by perturbing the phase image to be calibrated within a preset range; the phase perturbation within the preset range is determined based on historical phase errors, or based on the array phase of the actual optical phased array and laser source. The error determination module is used to input the phase image to be calibrated and the random perturbation phase image into a target convolutional neural network model to determine the target phase error of the phase to be calibrated. The target convolutional neural network model is trained on a preset convolutional neural network model based on a pair of far-field training images. The target convolutional neural network model includes a nonlinear mapping relationship between the far-field image and the phase. The far-field training image pair includes a far-field training image generated from an optical phased array and a far-field random perturbation phase image obtained by adding Gaussian noise to the electrical signal of the far-field training image. The module is also used to determine the predicted phase error based on the preset convolutional neural network model; and to determine the target convolutional neural network model based on a preset normalized evaluation function, according to the predicted phase error and the true compensation phase corresponding to the far-field training image pair. The network model; wherein, the determination of the target convolutional neural network model based on the predicted phase error and the true compensation phase corresponding to the far-field training image pair according to the preset normalized evaluation function includes: calculating the phase difference between the predicted phase error and the true compensation phase; when it is determined that the phase difference is a multiple of the first phase, inputting the far-field training image pair and the random phase distribution corresponding to the far-field training image pair into the preset convolutional neural network model for training until the phase difference is a multiple of the second phase; when it is determined that the phase difference is a multiple of the second phase, determining the training termination time in the preset convolutional neural network model, and determining the target convolutional neural network model after the training termination time is reached at the current time; the first phase is π, and the second phase is 2π; The calibration module is used to calibrate the initial emission electric field of the phase image to be calibrated based on the target phase error to obtain the target emission electric field; when it is determined that the far-field intensity of the target emission electric field meets the preset intensity threshold, a target phase image is generated based on the target emission electric field; the target phase image includes the image generated after calibrating the phase image to be calibrated.

8. An electronic device comprising a memory and a processor, characterized in that, The memory stores a computer program, and the processor is configured to run the computer program to perform the optical phased array phase calibration method according to any one of claims 1 to 6.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the steps of the optical phased array phase calibration method according to any one of claims 1 to 6.

Images