Synthetic aperture radar target identification method, image processing device, medium and equipment
By converting synthetic aperture radar images into light intensity distribution and optimizing phase distribution, combined with nonlinear processing, and using optical neural network modules to achieve high-precision, low-latency target recognition, the problems of low recognition rate of traditional methods and high energy consumption of deep learning are solved.
Patent Information
- Application Number
- CN202510523851.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-04-24
- Publication Date
- 2025-09-09
AI Technical Summary
Traditional synthetic aperture radar target recognition methods have low recognition rates in complex environments, and deep learning methods consume high energy on satellite-borne equipment, which cannot meet power consumption constraints.
By converting synthetic aperture radar images into light intensity distribution, optimizing phase distribution and performing phase modulation, combining nonlinear processing, extracting target features and classifying them, and using amplitude and phase spatial light modulator modules to realize nonlinear calculation of optical neural networks.
High-precision, low-latency synthetic aperture radar image target recognition is achieved, which improves the recognition rate and reduces energy consumption.
Smart Images
Figure CN120610265A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of synthetic aperture radar remote sensing information technology, and in particular to a synthetic aperture radar target recognition method, image processing device, medium and equipment. Background Art
[0002] In the field of target recognition, radar remote sensing equipment (such as synthetic aperture radar) has become an increasingly important technical tool. Although synthetic aperture radar images can provide high-resolution and multi-angle information, the complex scattering characteristics and coherent speckle noise make target feature extraction difficult.
[0003] Traditional synthetic aperture radar target recognition methods face a dual dilemma: 1) At the traditional algorithm level, methods based on Hu moments and scattering center matching are limited by the ability to express artificial features, and their recognition rates in complex environments are generally less than 75%; 2) At the deep learning level, although convolutional neural networks can increase the recognition rate of MSTAR datasets to 90%, the energy consumption of single-frame processing on GPU platforms exceeds 10J, which cannot meet the strict power consumption constraints of spaceborne equipment (usually required to be less than 1W). Summary of the Invention
[0004] Based on this, it is necessary to propose a synthetic aperture radar target recognition method, image processing device, medium and equipment to address the above problems.
[0005] A synthetic aperture radar target recognition method, the method comprising:
[0006] The synthetic aperture radar image is converted into a first light intensity distribution.
[0007] The initial phase distribution is optimized, and the first light intensity distribution is phase-modulated according to the optimized phase distribution to obtain a second light intensity distribution including light intensity and phase weights.
[0008] Nonlinear processing is performed on the second light intensity distribution to obtain a light intensity distribution map, target features are extracted from the light intensity distribution map, the target features are classified, and a target category is output.
[0009] The optimizing the phase distribution and performing phase modulation on the first light intensity distribution according to the optimized phase distribution to obtain a second light intensity distribution of a target feature specifically includes:
[0010] Set the maximum number of iterations, inertia weight, individual learning factor, and social learning factor.
[0011] The velocity of the particle is updated according to the inertia weight, the individual learning factor, and the social learning factor, the particle being a pixel, and the velocity being an adjustment amount of the phase distribution.
[0012] The position of the particle is updated according to the velocity of the particle, and the position is a phase distribution.
[0013] The speed and position of the particle are repeatedly updated until the maximum number of iterations is reached to obtain an optimized phase distribution.
[0014] Phase modulation is performed on the first light intensity distribution according to the optimized phase distribution to obtain a second light intensity distribution including light intensity and phase weights.
[0015] The updating of the particle velocity according to the inertia weight, the individual learning factor, and the social learning factor, where the particle is a pixel and the velocity is an adjustment amount of the phase distribution, specifically includes:
[0016] according to
[0017] vi,j=w·vi,j+c1·rand()·(pBesti,j-Positioni,j)+c2·rand()·(gBestj-Positioni,j) updates the velocity of the particle, where vi,j is the velocity of the i-th particle in the j-th dimension, w is the inertia weight, c1 is the individual learning factor, c2 is the social learning factor, pBesti,j is the historical best position of the i-th particle in the j-th dimension, gBestj is the global best position, and rand() is a random number between [0,1].
[0018] Wherein, the updating of the particle position according to the particle velocity is a phase distribution. Specifically includes:
[0019] according to Update the position of the particle, where position is the phase distribution and rand() is a random number between [0,1].
[0020] Wherein, after performing nonlinear processing on the second light intensity distribution to obtain a light intensity distribution map of a target feature and extracting the target feature from the light intensity distribution map, the method further includes:
[0021] The optimized phase distribution is evaluated based on the matching degree between the second light intensity distribution and the target feature, and the light intensity distribution is optimized again based on the evaluation result.
[0022] The converting of the synthetic aperture radar image into the first light intensity distribution specifically includes:
[0023] The parallel light is encoded according to the grayscale value of the synthetic aperture radar image to obtain a first light intensity distribution.
[0024] A synthetic aperture radar image processing device, characterized in that the device is used to execute the synthetic aperture radar target recognition method described above, and the device comprises:
[0025] an amplitude-type spatial light modulator module, configured to convert a synthetic aperture radar image into a first light intensity distribution;
[0026] a phase-type spatial light modulator module, configured to perform phase modulation on the first light intensity distribution according to the optimized phase distribution to obtain a second light intensity distribution of a target feature;
[0027] The photoelectric detector module is used to perform nonlinear processing on the second light intensity distribution to obtain a light intensity distribution diagram of the target feature.
[0028] The synthetic aperture radar image processing device further includes: a laser, a beam expander lens, a half-wave plate, a high-power reflector, a polarization beam splitter, and a non-polarization beam splitter.
[0029] The laser is used to emit planar light;
[0030] The first beam expander lens is used to expand and collimate the plane light;
[0031] The half-wave plate is used to adjust the polarization direction of plane light;
[0032] The high-power reflector is used to reflect the plane light after the polarization direction is adjusted;
[0033] The polarization beam splitter is used to transmit the plane light after the polarization direction is adjusted to the amplitude spatial light modulator module;
[0034] The non-polarization beam splitter is used to transmit the first light intensity distribution to the phase-type spatial light modulator module.
[0035] A computer-readable storage medium stores a computer program, which, when executed by a processor, causes the processor to perform the steps of the above method.
[0036] A computer device comprises a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the processor executes the steps of the method as described above.
[0037] The embodiments of the present invention have the following beneficial effects:
[0038] The present invention improves the modulation performance of the first light intensity distribution by adjusting the phase distribution, obtaining a second light intensity distribution containing light intensity and phase weights. Furthermore, a nonlinear response mechanism is employed to nonlinearly process the second light intensity distribution containing light intensity and phase weights to obtain a light intensity distribution map. Target features are extracted from the light intensity distribution map, classified, and the target category is output. By converting synthetic aperture radar images into light intensity distributions, high-precision, low-latency target recognition is achieved using synthetic aperture radar images. BRIEF DESCRIPTION OF THE DRAWINGS
[0039] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.
[0040] in:
[0041] Figure 1 A schematic flow chart of an embodiment of a synthetic aperture radar target recognition method provided by the present invention;
[0042] Figure 2 A schematic flow chart of another embodiment of the synthetic aperture radar target recognition method provided by the present invention;
[0043] Figure 3 A schematic diagram of the structure of an embodiment of a synthetic aperture radar image processing device provided by the present invention;
[0044] Figure 4 A schematic structural diagram of another embodiment of a synthetic aperture radar image processing device provided by the present invention;
[0045] Figure 5 A schematic structural diagram of an embodiment of the device provided by the present invention;
[0046] Figure 6 This is a schematic structural diagram of an embodiment of the medium provided by the present invention. DETAILED DESCRIPTION
[0047] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making any creative efforts shall fall within the scope of protection of the present invention.
[0048] like Figure 1 As shown, Figure 1 A flow chart of an embodiment of a synthetic aperture radar target recognition method provided by the present invention. A synthetic aperture radar target recognition method, the method comprising:
[0049] S101: Converting a synthetic aperture radar image into a first light intensity distribution.
[0050] Exemplarily, a synthetic aperture radar image to be processed is acquired in real time, and parallel light is encoded based on the grayscale values of the synthetic aperture radar image to obtain a first light intensity distribution. Specifically, a two-dimensional matrix of grayscale values of the target image is acquired. The parallel light is encoded based on the two-dimensional matrix of current grayscale values to obtain the first light intensity distribution.
[0051] S102: Optimizing the initial phase distribution, performing phase modulation on the first light intensity distribution according to the optimized phase distribution, and obtaining a second light intensity distribution including light intensity and phase weights.
[0052] For example, first, an initial phase distribution is set according to the frequency domain sparsity of the synthetic aperture radar image so that it can initially focus on the micro-Doppler characteristic area of the target.
[0053] Furthermore, the maximum number of iterations, inertia weight, individual learning factor, and social learning factor are set; the velocity of the particle is updated according to the inertia weight, individual learning factor, and social learning factor, where the particle is a pixel and the velocity is the adjustment amount of the phase distribution; the position of the particle is updated according to the velocity of the particle, where the position is the phase distribution; the velocity and position of the particle are repeatedly updated until the maximum number of iterations is reached, and the optimized phase distribution is obtained; the first light intensity distribution is phase modulated according to the optimized phase distribution to obtain a second light intensity distribution containing light intensity and phase weights.
[0054] S103: Perform nonlinear processing on the second light intensity distribution to obtain a light intensity distribution map, extract target features from the light intensity distribution map, classify and identify the target features, and output a target category.
[0055] Exemplarily, a nonlinear excitation function is applied to the second light intensity distribution, thereby implementing nonlinear processing in the optical system and obtaining a light intensity distribution map. Nonlinear excitation is a key step in simulating the nonlinear computation of optical neural networks, providing the optical neural network model with enhanced expressive power. Furthermore, target features are extracted from the light intensity distribution map, classified and identified using a support vector machine (SVM) or convolutional neural network (CNN) model, and the target category is output.
[0056] As can be seen from the above description, the present invention improves the modulation performance of the first light intensity distribution by adjusting the phase distribution, thereby obtaining a second light intensity distribution comprising light intensity and phase weights. Furthermore, a nonlinear response mechanism is employed to nonlinearly process the second light intensity distribution comprising light intensity and phase weights to obtain a light intensity distribution map. Target features are extracted from the light intensity distribution map, classified, and the target category is output. By converting synthetic aperture radar images into light intensity distributions, high-precision, low-latency target recognition is achieved within synthetic aperture radar images.
[0057] like Figure 2 As shown, Figure 2A flow chart of another embodiment of the synthetic aperture radar target recognition method provided by the present invention. A synthetic aperture radar target recognition method, the method comprising:
[0058] S201: Encode parallel light according to the grayscale value of the synthetic aperture radar image to obtain a first light intensity distribution.
[0059] It should be noted that step S201 Figure 1 This has been discussed in detail in the implementation scenario shown and will not be repeated here.
[0060] S202: Set the maximum number of iterations, inertia weight, individual learning factor, and social learning factor.
[0061] S203: Update the velocity of the particle according to the inertia weight, the individual learning factor, and the social learning factor. The particle is a pixel and the velocity is the adjustment amount of the phase distribution.
[0062] For example, before optimizing the SLM2 phase distribution, it is necessary to initialize relevant parameters according to the characteristics of the SAR image. These parameters include:
[0063] Input light field characteristics: Initialize the input light field parameters based on the resolution of the SAR image (e.g., 0.3m to 1m), frequency domain sparsity (Fourier transform characteristics of the SAR image), and micro-Doppler characteristics (motion modulation information of the target).
[0064] SLM2 parameters: Set the resolution (1920 × 1080) and refresh rate (200 Hz) of the SLM2 to ensure that it can dynamically reconstruct the light field to adapt to the characteristics of the SAR image.
[0065] At the same time, initialize the algorithm parameters: set the maximum number of iterations, inertia weight, individual learning factor, and social learning factor.
[0066] Furthermore, the particle velocity is updated according to the following formula:
[0067] vi,j=w·vi,j+c1·rand()·(pBesti,j-Positioni,j)+c2·rand()·(gBestj-Positioni,j);
[0068] Where vi,j is the velocity of the i-th particle in the j-th dimension, w is the inertia weight, c1 is the individual learning factor, c2 is the social learning factor, pBesti,j is the historical best position of the i-th particle in the j-th dimension, gBestj is the global best position, and rand() is a random number between [0,1].
[0069] S204: Update the position of the particle according to the velocity of the particle, and the position is a phase distribution.
[0070] For example, based on the updated velocity, an S-type or V-type transfer function is used to update the particle position, i.e., the phase distribution of SLM2. The particle position is updated according to the following formula:
[0071]
[0072] Among them, Position i,j is the position of the particle, the position is the phase distribution, and rand() is a random number between [0,1].
[0073] S205: Repeatedly update the particle velocity and position until the maximum number of iterations is reached to obtain the optimized phase distribution.
[0074] For example, the particle velocity and position are repeatedly updated until the maximum number of iterations is reached, the optimized phase distribution is obtained, and its light field effect is verified through experiments, and fine-tuning is performed if necessary to further improve the performance.
[0075] S206: Phase modulating the first light intensity distribution according to the optimized phase distribution to obtain a second light intensity distribution including light intensity and phase weights.
[0076] Exemplarily, the first light intensity distribution is phase modulated according to the optimized phase distribution to obtain a second light intensity distribution including light intensity and phase weights. The second light intensity distribution is nonlinearly processed to obtain a light intensity distribution map, target features are extracted from the light intensity distribution map, the target features are classified and identified, and the target category is output.
[0077] Furthermore, the optimized phase distribution is evaluated according to the matching degree between the second light intensity distribution and the target feature, and the light intensity distribution is optimized again according to the evaluation result.
[0078] S207: Perform nonlinear processing on the second light intensity distribution to obtain a light intensity distribution map, extract target features from the light intensity distribution map, classify and identify the target features, and output a target category.
[0079] like Figure 3 As shown, Figure 3 This is a schematic structural diagram of an embodiment of a synthetic aperture radar image processing device provided by the present invention. The device is used to perform the synthetic aperture radar target recognition method described above, and the device includes:
[0080] The amplitude spatial light modulator module 11 is configured to convert a synthetic aperture radar image into a first light intensity distribution.
[0081] Exemplarily, the amplitude-type spatial light modulator module 11 encodes the parallel light according to the grayscale value of the synthetic aperture radar image, obtains a first light intensity distribution, and transmits the first light intensity distribution to the phase-type spatial light modulator module 12 through an optical path.
[0082] The phase-type spatial light modulator module 12 is used to perform phase modulation on the first light intensity distribution according to the optimized phase distribution to obtain a second light intensity distribution with target characteristics.
[0083] Exemplarily, the phase-type spatial light modulator module 12 is composed of a liquid crystal array, receives the first light intensity distribution output by the amplitude-type spatial light modulator 11, phase-modulates the first light intensity distribution according to the currently optimized phase distribution, and obtains the second light intensity distribution of the target feature that currently includes light intensity and phase weights.
[0084] The photoelectric detector module 13 is used to perform nonlinear processing on the second light intensity distribution to obtain a light intensity distribution diagram of the target feature.
[0085] For example, after modulation by the phase-type spatial light modulator 12, the light field carries the current phase weight information. The photodetector module 13 performs nonlinear processing on the second light intensity distribution to obtain a light intensity distribution map of the target feature. By applying a nonlinear excitation function to the light field via the photodetector module 13, nonlinear processing in the optical system is achieved. Nonlinear excitation is a key step in simulating nonlinear computations in optical neural networks, providing enhanced expressive power for optical neural network models.
[0086] As can be seen from the above description, the present invention realizes phase and amplitude modulation of the first light intensity distribution and the second light intensity distribution by combining an amplitude-type spatial light modulation module and a phase-type spatial light modulation module. In addition, the integration of the amplitude-type spatial light modulator, the phase-type spatial light modulator and the detector constructs a compact optical neural network architecture with a simple system structure. The second light intensity distribution is processed nonlinearly by the photodetector module, thereby simulating the weights and nonlinear activation operations of the optical neural network to more efficiently complete the nonlinear mapping of complex data, significantly improving the modeling ability and reasoning accuracy of the optical neural network, and providing a foundation for achieving high-precision, low-latency synthetic aperture radar image target recognition.
[0087] like Figure 4 As shown, Figure 4 This is a schematic structural diagram of another embodiment of a synthetic aperture radar image processing device provided by the present invention. The synthetic aperture radar image processing device further includes: a laser 14, a beam expander lens 15, a half-wave plate 16, a high-power reflector 17, a polarizing beam splitter 18, a non-polarizing beam splitter 19, and a prism 20:
[0088] The laser 14 is used to emit planar light.
[0089] The beam expander lens 15 is used to expand and collimate the plane light.
[0090] The prism 20 is used to transmit plane light.
[0091] The half-wave plate 16 is used to adjust the polarization direction of the plane light.
[0092] The high-power reflector 17 is used to reflect the plane light after the polarization direction is adjusted.
[0093] The polarization beam splitter 18 is used to transmit the plane light with adjusted polarization direction to the amplitude-type spatial light modulator module.
[0094] The non-polarization beam splitter 19 is used to transmit the first light intensity distribution to the phase-type spatial light modulator module.
[0095] Exemplarily, the laser 14 emits highly coherent green plane light. The laser 14 is a 532nm laser. The plane light is expanded and collimated by the beam expander lens 15 to have a more uniform light intensity distribution and a smaller divergence angle. The plane light is transmitted to the half-wave plate 16 through the prism 20, and the polarization direction of the plane light is adjusted by the half-wave plate 16. Furthermore, the plane light after the polarization direction is adjusted is reflected to the polarization beam splitter 18 by the high-power reflector 17, and the plane light after the polarization direction is adjusted is transmitted to the amplitude-type spatial light modulator module 11 through the polarization beam splitter 18. The amplitude-type spatial light modulator module 11 encodes the plane light after the polarization direction is adjusted by the grayscale value of the synthetic aperture radar image, obtains a first light intensity distribution, and transmits the first light intensity distribution to the non-polarization beam splitter 19 through the optical path. The first light intensity distribution is transmitted to the phase-type spatial light modulator module 12 through the non-polarization beam splitter 19. The phase-type spatial light modulator module 12 receives the first light intensity distribution output by the amplitude-type spatial light modulator 11, optimizes the phase distribution of the phase-type spatial light modulator module 12, phase-modulates the first light intensity distribution according to the optimized phase distribution, and obtains the second light intensity distribution of the target feature that currently includes light intensity and phase weights.
[0096] From the above description, it can be seen that the present invention efficiently converts the synthetic aperture radar image into a first light intensity distribution through the amplitude-type spatial light modulator module, transmits it to the phase-type spatial light modulator module, and then performs complex transformations by the nonlinear activation module. Finally, the photoelectric detector module obtains the target classification result, thereby realizing high-precision and low-latency synthetic aperture radar image target recognition.
[0097] like Figure 5 As shown, Figure 5The device 30 includes a memory 31 and a processor 32. The memory 31 stores a computer program, and the processor 32 executes the computer program when working to achieve the following. Figure 1 and Figure 2 The method shown.
[0098] The specific technical details of a synthetic aperture radar target recognition method implemented when the above-mentioned device 30 executes the computer program have been discussed in detail in the aforementioned method steps, so they will not be repeated here.
[0099] like Figure 6 As shown, Figure 6 The structure diagram of an embodiment of the medium provided by the present invention. The medium 40 stores at least one computer program 41, which is executed by the processor 32 to implement the following Figure 1 and Figure 2 In one embodiment, the medium 40 may be a memory chip, a hard disk, a mobile hard disk, a USB flash drive, an optical disk, or other readable and writable storage tools, or a server.
[0100] The foregoing description of specific embodiments of the present disclosure is intended to illustrate a method for performing a multi-tasking process. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recited in the claims may be performed in an order different from that described in the embodiments and still achieve the desired results. Furthermore, the processes depicted in the accompanying drawings do not necessarily require the specific order shown or the sequential order shown to achieve the desired results. In certain embodiments, multitasking and parallel processing are also possible or may be advantageous.
[0101] The various embodiments in this specification are described in a progressive manner. Similar portions between the various embodiments can be referenced to each other, and each embodiment focuses on the differences from the other embodiments. In particular, the device, apparatus, and non-volatile computer-readable storage medium embodiments are generally similar to the method embodiments, so their descriptions are relatively simplified. For relevant details, refer to the descriptions of the method embodiments.
[0102] The apparatus, device, non-volatile computer-readable storage medium and method provided in the embodiments of this specification correspond to each other. Therefore, the apparatus, device, and non-volatile computer storage medium also have similar beneficial technical effects as the corresponding method. Since the beneficial technical effects of the method have been described in detail above, the beneficial technical effects of the corresponding apparatus, device, and non-volatile computer storage medium will not be repeated here.
[0103] The systems, devices, modules, or units described in the above embodiments may be implemented by computer chips or entities, or by products having certain functions. A typical implementation device is a computer. Specifically, the computer may be, for example, a personal computer, a laptop computer, a cellular phone, a camera phone, a smartphone, a personal digital assistant, a media player, a navigation device, an email device, a game console, a tablet computer, a wearable device, or a combination of any of these devices.
[0104] For the convenience of description, when describing the above device, various units are divided into functions and described separately. Of course, when implementing this specification, the functions of each unit can be implemented in the same one or more software and / or hardware. It should be understood by those skilled in the art that this specification embodiment can be provided as a method, system, or computer program product. Therefore, this specification embodiment can adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, this specification embodiment can adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.
[0105] Computer-readable media include permanent and non-permanent, removable and non-removable media that can be implemented by any method or technology to store information. The information can be computer-readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disc read-only memory (CD-ROM), digital versatile disc (DVD) or other optical storage, magnetic cassettes, magnetic disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer-readable media does not include transitory media such as modulated data signals and carrier waves.
[0106] The various embodiments in this specification are described in a progressive manner. Similar parts between the various embodiments can be referred to in conjunction with each other. Each embodiment focuses on the differences between the other embodiments. In particular, the system embodiments are generally similar to the method embodiments, so the description is relatively simple. For relevant parts, refer to the description of the method embodiments.
[0107] The above disclosure is merely a preferred embodiment of the present invention and certainly cannot be used to limit the scope of the present invention. Therefore, equivalent changes made according to the claims of the present invention are still within the scope of the present invention.
Claims
1. A synthetic aperture radar target recognition method, characterized in that: The method comprises: converting a synthetic aperture radar image into a first light intensity distribution; Optimizing the initial phase distribution, performing phase modulation on the first light intensity distribution according to the optimized phase distribution, and obtaining a second light intensity distribution including light intensity and phase weights; Nonlinear processing is performed on the second light intensity distribution to obtain a light intensity distribution map, target features are extracted from the light intensity distribution map, the target features are classified, and a target category is output.
2. The synthetic aperture radar target recognition method according to claim 1, characterized in that: Optimizing the phase distribution, performing phase modulation on the first light intensity distribution according to the optimized phase distribution, and obtaining a second light intensity distribution of a target feature specifically includes: Set the maximum number of iterations, inertia weight, individual learning factor, and social learning factor; updating the velocity of the particle according to the inertia weight, the individual learning factor, and the social learning factor, wherein the particle is a pixel and the velocity is an adjustment amount of the phase distribution; updating the position of the particle according to the velocity of the particle, wherein the position is a phase distribution; Repeatedly updating the velocity and position of the particle until the maximum number of iterations is reached to obtain an optimized phase distribution; Phase modulation is performed on the first light intensity distribution according to the optimized phase distribution to obtain a second light intensity distribution including light intensity and phase weights.
3. The synthetic aperture radar target recognition method according to claim 2, characterized in that: The updating of the particle velocity according to the inertia weight, the individual learning factor, and the social learning factor, wherein the particle is a pixel and the velocity is an adjustment amount of the phase distribution, specifically includes: Update the velocity of the particle according to vi,j = w·vi,j+c1·rand()·(pBesti,j-Positioni,j)+c2·rand()·(gBestj-Positioni,j), where vi,j is the velocity of the i-th particle in the j-th dimension, w is the inertia weight, c1 is the individual learning factor, c2 is the social learning factor, pBesti,j is the historical best position of the i-th particle in the j-th dimension, gBestj is the global best position, and rand() is a random number between [0,1].
4. The synthetic aperture radar target recognition method according to claim 3, characterized in that: The updating of the particle position according to the particle velocity, wherein the position is a phase distribution, specifically includes: according to Update the position of the particle, where position is the phase distribution and rand() is a random number between [0,1].
5. The synthetic aperture radar target recognition method according to claim 1, characterized in that: After performing nonlinear processing on the second light intensity distribution to obtain a light intensity distribution map of a target feature, and extracting the target feature from the light intensity distribution map, the method further includes: The optimized phase distribution is evaluated based on the matching degree between the second light intensity distribution and the target feature, and the light intensity distribution is optimized again based on the evaluation result.
6. The synthetic aperture radar target recognition method according to claim 1, characterized in that: The converting the synthetic aperture radar image into the first light intensity distribution specifically includes: The parallel light is encoded according to the grayscale value of the synthetic aperture radar image to obtain a first light intensity distribution.
7. A synthetic aperture radar image processing device, characterized in that: The device is used to execute the synthetic aperture radar target recognition method according to any one of claims 1 to 6, and the device includes: an amplitude-type spatial light modulator module, configured to convert a synthetic aperture radar image into a first light intensity distribution; a phase-type spatial light modulator module, configured to perform phase modulation on the first light intensity distribution according to the optimized phase distribution to obtain a second light intensity distribution of a target feature; The photoelectric detector module is used to perform nonlinear processing on the second light intensity distribution to obtain a light intensity distribution diagram of the target feature.
8. The synthetic aperture radar image processing device according to claim 7, characterized in that: The synthetic aperture radar image processing device further includes: a laser, a beam expanding lens, a half-wave plate, a high-power reflector, a polarization beam splitter, and a non-polarization beam splitter: The laser is used to emit planar light; The first beam expanding lens is used to expand and collimate the plane light; The half-wave plate is used to adjust the polarization direction of plane light; The high-power reflector is used to reflect the plane light after the polarization direction is adjusted; The polarization beam splitter is used to transmit the plane light after the polarization direction is adjusted to the amplitude spatial light modulator module; The non-polarization beam splitter is used to transmit the first light intensity distribution to the phase-type spatial light modulator module.
9. A computer-readable storage medium storing a computer program, wherein when the computer program is executed by a processor, the processor is caused to perform the steps of the method according to any one of claims 1 to 6.
10. A computer device comprising a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the processor is caused to perform the steps of the method according to any one of claims 1 to 6.