Target image classification method and system based on complex background, medium and equipment

Abstract

The embodiment of the invention discloses a target image classification method based on a complex background. The method comprises the following steps: acquiring continuous and stable monochromatic p polarized light; collecting a target image, loading the target image to an amplitude modulation array, modulating the monochromatic p polarized light by using the amplitude modulation array loaded with the target image, and obtaining a space time domain optical signal of the target image containing a complex background; converting the spatial time domain optical signal of the target image containing the complex background into a spatial frequency domain signal, wherein the spatial frequency domain signal is a spatial frequency distribution diagram containing the spatial information of the target image; background information of the spatial frequency domain signal is removed, and the spatial frequency domain signal with the background removed is obtained; converting the background-removed spatial frequency domain signal into a background-removed spatial time domain optical signal, wherein the background-removed spatial time domain optical signal is a spatial intensity distribution diagram containing target graphic spatial information; and target images corresponding to the space time domain optical signals after the background is removed are classified.

Description

Technical Field

[0001] The present invention relates to the technical field of image processing and recognition, and in particular to a method, system, medium and equipment for classifying target images based on complex backgrounds. Background Art

[0002] The traditional von Neumann architecture has long dominated the computing field. However, with the continuous advancement of technology and the growing demand, the limitations of this architecture have gradually become apparent. Due to the design principle of separation of storage and computing, the von Neumann architecture faces the dual challenges of power consumption and memory, which limits the improvement of computing performance and gradually slows down the growth rate of Moore's Law. In order to meet these challenges, researchers are constantly exploring new computing technologies and architectures in order to break through the limits of traditional microelectronic computing.

[0003] As an emerging computing technology, Diffractive Optical Neural Networks (DONN) has shown broad application prospects in the fields of high-performance computing and artificial intelligence due to its advantages of ultra-high speed, large bandwidth, and multi-dimensionality. DONN fully integrates advanced technologies such as high-speed optical communication, optical interconnection, optical integration, and silicon-based optoelectronics, providing a highly competitive solution to break through the bottleneck of traditional microelectronic computing in the post-Moore era.

[0004] However, despite the many advantages of DONN, it still faces some challenges in practical applications. At present, when using DONN for target recognition, it is mostly used to identify simple targets that have been processed in advance. When facing targets in complex backgrounds, DONN often cannot achieve accurate recognition, which greatly reduces its practicality in real-world scenarios. Therefore, how to improve DONN's target recognition ability in complex backgrounds has become a problem that needs to be solved urgently. Summary of the invention

[0005] Based on this, it is necessary to propose a target image classification method based on complex background to address the above problems.

[0006] A target image classification method based on complex background, the method comprising the following steps:

[0007] Obtain continuous and stable monochromatic p-polarized light;

[0008] Acquire a target image, and load the target image into an amplitude modulation array, and modulate the monochromatic p-polarized light using the amplitude modulation array after the target image is loaded, so as to obtain a spatial time-domain optical signal of the target image containing a complex background;

[0009] Converting the spatial time domain optical signal of the target image containing the complex background into a spatial frequency domain signal, wherein the spatial frequency domain signal is a spatial frequency distribution diagram containing the spatial information of the target image;

[0010] Removing background information of the spatial frequency domain signal to obtain a spatial frequency domain signal after the background is removed;

[0011] The spatial frequency domain signal after background removal is converted into a spatial time domain light signal after background removal, wherein the spatial time domain light signal after background removal is a spatial intensity distribution diagram containing spatial information of the target graphic; and the target image corresponding to the spatial time domain light signal after background removal is classified.

[0012] In the above solution, the step of obtaining continuous and stable monochromatic p-polarized light specifically includes:

[0013] The collection laser emits a continuous and stable monochromatic linearly polarized laser beam;

[0014] Passing the monochromatic linearly polarized laser beam through a beam expanding lens pair and adjusting the lens pair to obtain a parallel beam after beam expansion;

[0015] Performing polarization modulation on the expanded parallel light beam according to a half-wave plate;

[0016] The angle of the half-wave plate is controlled to obtain continuous and stable monochromatic p-polarized light.

[0017] In the above solution, the step of acquiring a target image and loading the target image into an amplitude modulation array further comprises: performing polarization filtering on the continuous and stable monochromatic p-polarized light through a polarization beam splitter prism PBS.

[0018] In the above scheme, the monochromatic p-polarized light is modulated by the amplitude modulation array after the target image is loaded to obtain the spatial-temporal light signal of the target image containing a complex background, which also includes: distributing the light field intensity of the reflected light according to the grayscale value of the loaded grayscale image to obtain the spatial-temporal light signal of the target image containing a complex background.

[0019] In the above solution, converting the spatial time domain optical signal of the target image containing a complex background into a spatial frequency domain signal specifically includes:

[0020] The amplitude modulation array of the reflective amplitude-type spatial light modulator SLM receives the monochromatic p-polarized light reflected by the first polarization beam splitter prism PBS, and the monochromatic p-polarized light is reflected as monochromatic s-polarized light after amplitude modulation;

[0021] The first polarization beam splitter prism PBS reflects the monochromatic s-polarized light in a direction perpendicular to the incident light path;

[0022] The reflected light beam passes through the first polarization beam splitter prism PBS from the amplitude modulation array, and the focal length f of the lens is determined according to the distance of the light beam in its propagation path when it reaches the Fourier lens;

[0023] When the spatial time domain optical signal of the target image with a complex background passes through the Fourier lens, the Fourier lens converts it into the spatial frequency domain to obtain a spatial frequency domain signal.

[0024] In the above solution, converting the spatial time domain optical signal of the target image containing a complex background into a spatial frequency domain signal specifically includes:

[0025] Perform a two-dimensional Fourier transform on the modulated s-polarized light to convert the spatial information of the target image into frequency information;

[0026] According to the spatial frequency distribution diagram formed on the rear focal plane of the Fourier lens, the spatial frequency domain signal of the target image containing a complex background is determined by the signal after Fourier transformation.

[0027] In the above solution, filtering and nonlinearly activating the spatial frequency domain signal to obtain the processed spatial frequency domain signal specifically includes:

[0028] Performing filtering processing on the spatial frequency domain signal according to a first diffractive optical neural network DONN;

[0029] The spatial frequency domain signal after filtering is nonlinearly activated according to the transmissive phase-type spatial light modulator SLM to obtain the processed spatial frequency domain signal.

[0030] The present application also proposes a target image classification system based on complex background, the system comprising: a polarization acquisition unit, a time domain signal acquisition unit, a signal processing unit and a classification unit;

[0031] The polarization acquisition unit is used to acquire continuous and stable monochromatic p polarization;

[0032] The time domain signal acquisition unit is used to collect a target image, load the target image into an amplitude modulation array, and modulate the monochromatic p-polarized light using the amplitude modulation array loaded with the target image to acquire a spatial time domain optical signal of the target image containing a complex background;

[0033] The signal processing unit is used to convert the spatial time domain light signal of the target image containing a complex background into a spatial frequency domain signal, wherein the spatial frequency domain signal is a spatial frequency distribution diagram containing spatial information of the target image; remove the background information of the spatial frequency domain signal to obtain the spatial frequency domain signal after the background is removed; convert the spatial frequency domain signal after the background is removed into a spatial time domain light signal after the background is removed, wherein the spatial time domain light signal after the background is removed is a spatial intensity distribution diagram containing spatial information of the target graphic; and the classification unit is used to classify the target image corresponding to the spatial time domain light signal after the background is removed.

[0034] The present application also proposes a readable storage medium storing a computer program, wherein when the computer program is executed by a processor, the processor executes the following steps:

[0035] Obtain continuous and stable monochromatic p-polarized light;

[0036] Acquire a target image, and load the target image into an amplitude modulation array, and modulate the monochromatic p-polarized light using the amplitude modulation array after the target image is loaded, so as to obtain a spatial time-domain optical signal of the target image containing a complex background;

[0037] Converting the spatial time domain optical signal of the target image containing the complex background into a spatial frequency domain signal, wherein the spatial frequency domain signal is a spatial frequency distribution diagram containing the spatial information of the target image;

[0038] Removing background information of the spatial frequency domain signal to obtain a spatial frequency domain signal after the background is removed;

[0039] The spatial frequency domain signal after background removal is converted into a spatial time domain light signal after background removal, wherein the spatial time domain light signal after background removal is a spatial intensity distribution diagram containing spatial information of the target graphic; and the target image corresponding to the spatial time domain light signal after background removal is classified.

[0040] The present application also proposes a computer device, comprising a memory and a processor, wherein the memory stores a computer program, and the processor executes the following steps of the computer program:

[0041] Obtain continuous and stable monochromatic p-polarized light;

[0042] Acquire a target image, and load the target image into an amplitude modulation array, and modulate the monochromatic p-polarized light using the amplitude modulation array after the target image is loaded, so as to obtain a spatial time-domain optical signal of the target image containing a complex background;

[0043] Converting the spatial time domain optical signal of the target image containing the complex background into a spatial frequency domain signal, wherein the spatial frequency domain signal is a spatial frequency distribution diagram containing the spatial information of the target image;

[0044] Removing background information of the spatial frequency domain signal to obtain a spatial frequency domain signal after the background is removed;

[0045] The spatial frequency domain signal after background removal is converted into a spatial time domain light signal after background removal, wherein the spatial time domain light signal after background removal is a spatial intensity distribution diagram containing spatial information of the target graphic; and the target image corresponding to the spatial time domain light signal after background removal is classified.

[0046] The use of the embodiment of the present invention has the following beneficial effects: first obtain continuous and stable monochromatic p-polarized light; collect the target image, and load the target image into the amplitude modulation array, use the amplitude modulation array after loading the target image to modulate the monochromatic p-polarized light, and obtain the spatial time-domain light signal of the target image containing a complex background; convert the spatial time-domain light signal of the target image containing a complex background into a spatial frequency domain signal, and the spatial frequency domain signal is a spatial frequency distribution diagram containing the spatial information of the target image; remove the background information of the spatial frequency domain signal, and obtain the spatial frequency domain signal after the background is removed; convert the spatial frequency domain signal after the background is removed into the spatial time-domain light signal after the background is removed, and the spatial time-domain light signal after the background is removed is a spatial intensity distribution diagram containing the spatial information of the target graphic; classify the target image corresponding to the spatial time-domain light signal after the background is removed. The present invention can extract the target image from the complex background and accurately classify it by processing the optical signal. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. 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 creative work.

[0048] in:

[0049] Figure 1 A schematic diagram of a target image classification method based on complex background in one embodiment;

[0050] Figure 2 Schematic diagram of the structure of a target image classification system based on complex background in one embodiment.

[0051] Description of Reference Numerals

[0052] 1: Laser; 2: Beam expansion lens; 3: HWP; 4: PBS; 5: Reflective amplitude-type SLM; 6: Target acquisition detector; 7: Processing system; 8: Fourier lens; 9: First DONN; 10: Transmissive phase-type SLM; 11: Anti-Fourier lens; 12: Second DONN; 13: Classification and recognition detector. DETAILED DESCRIPTION

[0053] The technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments; based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0054] In the following description, a large number of specific details are given in order to provide a more thorough understanding of the present invention; however, it is obvious to those skilled in the art that the present invention can be implemented without one or more of these details; in other examples, in order to avoid confusion with the present invention, some technical features known in the art are not described, and it should be understood that the present invention can be implemented in different forms and should not be interpreted as limited to the embodiments presented here; on the contrary, providing these embodiments will make the disclosure thorough and complete and fully convey the scope of the present invention to those skilled in the art.

[0055] The purpose of the terms used herein is only to describe specific embodiments and is not intended to be limiting of the present invention. When used herein, the singular forms "a", "an" and "said / the" are also intended to include the plural forms, unless the context clearly indicates otherwise, and it should also be understood that the terms "consisting of" and / or "comprising", when used in this specification, determine the presence of the features, integers, steps, operations, elements and / or parts, but do not exclude the presence or addition of one or more other features, integers, steps, operations, elements, parts and / or groups. When used herein, the term "and / or" includes any and all combinations of the relevant listed items.

[0056] In order to solve this problem, this application proposes a new method, that is, to perform Fourier transform on the optical signal acquired in real time in the spatial domain through the 4f system, convert it into an optical signal in the spatial frequency domain, and then use DONN and a spatial light modulator (Spatial Light Modulator, SLM) to filter and nonlinearly activate the optical signal in the spatial frequency domain. This method can effectively filter out complex backgrounds and retain the target to be identified, thereby improving the target recognition ability of DONN in complex backgrounds. Finally, another DONN is used to identify the target image after filtering the background to achieve accurate target classification. This method not only improves the practicality of DONN, but also provides a broader space for its application in daily life.

[0057] In order to thoroughly understand the present invention, a detailed structure will be presented in the following description to illustrate the technical solution proposed by the present invention; optional embodiments of the present invention are described in detail as follows, but in addition to these detailed descriptions, the present invention may also have other implementation methods.

[0058] like Figure 1 As shown, in one embodiment, a target image classification method based on a complex background is provided, and the target image classification method based on a complex background includes steps S101 to S106, which are described in detail as follows:

[0059] S101, obtaining continuous and stable monochromatic p-polarized light;

[0060] In some embodiments, obtaining continuous and stable monochromatic p-polarized light specifically includes:

[0061] The collection laser emits a continuous and stable monochromatic linearly polarized laser beam;

[0062] Pass a monochromatic linearly polarized laser beam through a beam expanding lens pair and adjust the lens pair to obtain a parallel beam after beam expansion;

[0063] Polarization modulation is performed on the expanded parallel light beam according to the half-wave plate;

[0064] Control the angle of the half-wave plate to obtain continuous and stable monochromatic p-polarized light.

[0065] Preferably, the laser emits a continuous and stable monochromatic linearly polarized laser beam, which is expanded and adjusted into a parallel beam by a beam expander lens. At this time, the incident light wave is a plane wave, and its complex amplitude U(r) is expressed as:

[0066] U(r)=u 0 e ikz

[0067] Among them, u 0is the real part of the complex amplitude, which determines the light intensity, and the phase function It means that the phase only changes with the z direction and has nothing to do with x and y. The wave vector k represents the propagation direction of the wave, and the expression is k = 2π / λ.

[0068] Since the normal operation of SLM requires p-polarized light (the component of the electric vector vibrating parallel to the incident plane is called the p component, and if the linear polarized light has only the p component, it is called p-polarized light) to be incident, it is also necessary to perform polarization modulation on the expanded light beam through HWP to obtain continuous and stable monochromatic p-polarized light.

[0069] The polarization modulation process is to rotate the half-wave plate so that its fast axis direction forms an angle θ with the original polarization direction. Since the half-wave plate adds a phase delay π to the slow axis direction component, the linearly polarized light remains linearly polarized light after passing through the half-wave plate and the orientation of the vibration plane is rotated by 2θ compared to the incident light. Therefore, the polarization direction of the linearly polarized light can be adjusted to become p-polarized light by simply changing the size of the angle θ.

[0070] S102, collecting a target image, and loading the target image into an amplitude modulation array, using the amplitude modulation array loaded with the target image to modulate the monochromatic p-polarized light, to obtain a spatial time-domain optical signal of the target image containing a complex background;

[0071] In some embodiments, acquiring the target image and loading the target image into the amplitude modulation array further comprises: performing polarization filtering on the continuous and stable monochromatic p-polarized light through a polarization beam splitter prism PBS.

[0072] Specifically, the PBS performs polarization filtering on the outgoing light beam modulated by the HWP, which can transmit p-polarized light and reflect s-polarized light, so that only p-polarized light is incident on the amplitude modulation array of the reflective amplitude-type SLM. At the same time, according to actual use requirements, the target grayscale image containing a complex background acquired in real time by the target acquisition detector is loaded on the amplitude modulation array of the reflective amplitude-type SLM through the processing system, so that the light field intensity of the reflected light is distributed according to the grayscale value of the loaded grayscale image;

[0073] The working principle of PBS is to select a suitable adhesive (n e <n<n o ) Put two negative uniaxial crystals (n e <n o ) are glued together along the inclined plane, and the optical axes of the two s crystals are parallel to the right-angled plane through which light passes and the incident plane. At this time, the e light is p-polarized light, and the o light is s-polarized light (the component of the electric vector vibrating perpendicular to the incident plane is called the s component. If a linear polarized light has only the s component, it is called s-polarized light);

[0074] The working principle of amplitude-type SLM is that when no electricity is applied to the electrode, the arrangement of liquid crystal gradually twists 90° along the light incident surface, and the natural light passes through the polarizer p1 (parallel to the incident plane direction), and the polarized light with the vibration direction parallel to the incident plane is obtained, that is, p-polarized light. After the light beam is incident on the light valve, it is reflected by the dielectric reflector and then emitted through the light valve. The polarization direction of the emitted light does not change, so no light passes through the orthogonal analyzer p 2 (perpendicular to the incident surface). When the electrodes of the liquid crystal light valve are powered on, the arrangement of the liquid crystal is twisted to 45°, and the vibration direction of the outgoing light reflected by the dielectric reflector is perpendicular to the incident surface. At this time, the outgoing light can all pass through the polarizer p 2 ;

[0075] According to actual use requirements, the grayscale image of the target acquired by the target acquisition detector is loaded on the amplitude modulation array through the processing module. The amplitude modulation array changes the resistance of the photosensitive layer according to the grayscale value of the pixel point written into the image, so that an electric field that varies with space is formed on the liquid crystal layer. The change in the distribution of the electric field applied to the liquid crystal layer changes the distortion of the liquid crystal arrangement, so that the vibration direction of the polarized light emitted through the light valve changes, and the intensity of the light emitted through the analyzer is distributed in space, thereby achieving the purpose of spatial light modulation. At this time, the reflected light field can be expressed as:

[0076]

[0077] Among them, u(x,y) is the gray value of the pixel at (x,y) in the image being written, and δ(x 0 -x,y 0 -y) is a pulse function. The light field at this time can be regarded as a linear combination of several point light sources, and its light field complex amplitude U(x 0 ,y 0 ) is distributed according to the grayscale u(x,y) of the loaded image.

[0078] Then the complex amplitude of the reflected light field of the reflective amplitude-type SLM is:

[0079] U(x 0 ,y 0 )=A(x 0 ,y 0 )r(x 0 ,y 0 )

[0080] Among them, A(x 0 ,y 0 ) is the light wave irradiated onto the amplitude modulation array, r(x 0 ,y 0 ) is the amplitude reflection coefficient, i.e., the grayscale distribution of the target image with complex background after horizontal inversion loaded on the amplitude modulation array;

[0081] Since the light source irradiating the amplitude modulation array is a plane light, its amplitude is assumed to be 1, then the complex amplitude of the reflected light field of the reflective amplitude type is:

[0082] U(x 0 ,y 0 )=r(x 0 ,y 0 )

[0083] Among them, r(x 0 ,y 0 ) is the amplitude reflection coefficient.

[0084] In some embodiments, the monochromatic p-polarized light is modulated by an amplitude modulation array after the target image is loaded to obtain the spatial-temporal light signal of the target image containing a complex background, which also includes: distributing the light field intensity of the reflected light according to the grayscale value of the loaded grayscale image to obtain the spatial-temporal light signal of the target image containing a complex background.

[0085] S103, converting the spatial time domain optical signal of the target image containing a complex background into a spatial frequency domain signal, wherein the spatial frequency domain signal is a spatial frequency distribution diagram containing spatial information of the target image;

[0086] In some embodiments, converting the spatial time domain optical signal of the target image containing a complex background into a spatial frequency domain signal specifically includes:

[0087] The amplitude modulation array of the reflective amplitude-type spatial light modulator SLM receives the monochromatic p-polarized light reflected by the first polarization beam splitter prism PBS, and the monochromatic p-polarized light is reflected as monochromatic s-polarized light after amplitude modulation;

[0088] The monochromatic s-polarized light is reflected in a direction perpendicular to the incident light path by the first polarization beam splitter prism PBS;

[0089] The reflected light beam passes through the first polarization beam splitter prism PBS from the amplitude modulation array, and the focal length f of the lens is determined according to the distance of the light beam in its propagation path when it reaches the Fourier lens;

[0090] When the spatial-time domain optical signal of the target image with a complex background passes through the Fourier lens, the Fourier lens converts it into the spatial-frequency domain to obtain the spatial-frequency domain signal.

[0091] Specifically, the incident p-polarized light is reflected as s-polarized light after being amplitude modulated by the reflective amplitude-type SLM, and is reflected by the first PBS in a direction perpendicular to the incident light path. The distance from the image of the reflected light beam from the amplitude modulation array through the PBS to the Fourier lens is exactly the focal length f of the Fourier lens.

[0092] When a light beam passes through multiple optical elements or optical systems, the image of the previous element or system is the object of the next element or system. When the light field reflected by the amplitude modulation array is reflected by the PBS, it is equivalent to passing through a parallel plate. At this time, the image of the target is located in front of the amplitude modulation array, and the distance from the amplitude modulation array is:

[0093]

[0094] Wherein, L is the side length of PBS, n is the refractive index of PBS, and at this time, the distance from the image of the reflected light field through PBS to the Fourier lens should be the focal length f of the Fourier lens to form a 4f system;

[0095] According to the above formula, the distance l from the PBS surface to the Fourier lens can be obtained 1 , PBS side length L, distance from amplitude modulation array to PBS surface l 2 And the relationship between the focal length f of the Fourier lens:

[0096] l 1 + L + Δl = f

[0097] Substituting Δl into the above formula, we can get

[0098]

[0099] When the above relationship is satisfied, a 4f system is formed;

[0100] The reflected light field is reflected by the PBS and then irradiated by the Fourier lens onto the front surface of the first DONN, where Fresnel diffraction occurs. The complex amplitude on the front surface of the first DONN is obtained from the Fresnel diffraction formula:

[0101] U(ξ,η)=c·FT{r(x 0 ,y 0 )}=c·R(ξ,η)

[0102] Among them, c is a complex constant, ξ=x / λf, η=y / λf, which correspond to the positions of x and y in the spatial frequency domain respectively. This transformation is called the standard Fourier transform. At this time, the target spatial domain optical signal containing complex background is converted to the spatial frequency domain.

[0103] In some embodiments, converting the spatial time domain optical signal of the target image containing a complex background into a spatial frequency domain signal specifically includes:

[0104] Perform a two-dimensional Fourier transform on the modulated s-polarized light to convert the spatial information of the target image into frequency information;

[0105] According to the spatial frequency distribution diagram formed on the focal plane, the spatial frequency domain signal of the target image containing a complex background is determined by the signal after Fourier transformation.

[0106] S104, removing background information of the spatial frequency domain signal to obtain a spatial frequency domain signal after the background is removed;

[0107] In some embodiments, removing background information of the spatial frequency domain signal to obtain the spatial frequency domain signal after the background is removed specifically includes:

[0108] Performing filtering processing on the spatial frequency domain signal according to the first diffractive optical neural network DONN;

[0109] The spatial frequency domain signal after filtering is nonlinearly activated according to the transmissive phase-type spatial light modulator SLM to obtain the processed spatial frequency domain signal.

[0110] Specifically, the optical signal converted into the spatial frequency domain will be filtered by the first DONN. The parameters of the first DONN are obtained by training on a computer using a deep learning method according to a specific filtering task data set. Assuming that the filtering function of the first DONN in the spatial frequency domain is H(ξ,η), the complex amplitude of the light field after filtering by the first DONN is:

[0111] U′(ξ,η)=U(ξ,η)H(ξ,η)

[0112] After filtering, part of the optical signal containing background information will be removed;

[0113] Specifically, the above H(ξ,η) can be composed of the superposition of the vortex phase function random matrix phase function and the Fresnel lens phase function. The vortex phase function modulates the light field so that its wavefront phase is spirally distributed. At this time, the light beam carries a certain orbital angular momentum, and the light intensity also presents a circular distribution. For the target light signal converted to Fourier space, part of the fundamental frequency information located in the center of the spatial frequency domain is filtered out by the vortex phase function, so that the final output only retains the medium and high frequency information.

[0114] Among them, the random matrix phase function is randomly distributed in two-dimensional space, ranging from 0 to 2π, and is used to accurately modulate the spatial frequency domain information to filter out the complex background so as to retain only the medium and high frequency information of the target, which is displayed as the target shape in the spatial domain. The Fresnel lens phase function is used to simulate the Fresnel lens, which is used to focus or diverge the light beam. Since part of the energy of the light field is diffracted outside the field of view of the system when passing through the DONN, resulting in a reduction in the final imaging intensity, it is necessary to add the Fresnel lens phase function in the first DONN to converge its energy to ensure the final processing effect.

[0115] S105, converting the spatial frequency domain signal after background removal into a spatial time domain optical signal after background removal, wherein the spatial time domain optical signal after background removal is a spatial intensity distribution diagram containing spatial information of the target graphic;

[0116] Among them, after the nonlinear activation of the transmissive phase-type SLM, the optical signal containing the target information will be enhanced and the remaining optical signal containing the background information will be weakened. At this time, the optical signal almost only contains the target information. The basic principle of nonlinear activation is to set the gamma correction curve of the transmissive phase-type SLM to a pre-selected nonlinear activation function Φ(ξ,η)=e f(ξ,η) , when the light field undergoes nonlinear activation, its complex amplitude can be expressed as:

[0117] U′(ξ,η)=U(ξ,η)H(ξ,η)Φ(ξ,η)=c·e f(ξ,η) R(ξ,η)H(ξ,η)

[0118] Among them, c is a complex constant and H(ξ,η) is the filter function.

[0119] The distances from the rear surface of the transmissive phase-type SLM and the front surface of the second DONN to the anti-Fourier lens are both the focal length f of the anti-Fourier lens. At this time, the optical signal in the spatial frequency domain will be Fourier transformed again. If the coordinates are opposite, it is equivalent to the optical signal after nonlinear activation being inversely Fourier transformed and transferred to the spatial domain. At this time, the optical signal in the spatial domain almost only has the target information, and its optical wave field can be expressed as:

[0120] U i (x i ,y i )=FT -1 {c·e f(ξ,η) R(ξ,η)H(ξ,η)}

[0121] Among them, c is a complex constant and H(ξ,η) is the filter function.

[0122] S106, classifying the target image corresponding to the spatial-temporal optical signal after the background is removed.

[0123] Specifically, the second DONN classifies the spatial domain light signal after removing the background. Its parameters are trained on a computer using deep learning methods based on a specific classification task data set. The classification and recognition detector acquires the classified light signal and transmits it to the processing system for recognition, thereby realizing real-time classification and recognition processing of targets under complex backgrounds.

[0124] like Figure 2 As shown, the present application also proposes a target image classification system based on complex background, the system comprising:

[0125] Laser, used to provide a continuous and stable light source for the entire system. The light it emits is monochromatic and linearly polarized light with a fixed polarization direction;

[0126] The beam expander lens is used to diverge the monochromatic beam emitted by the laser and collimate and shape the beam so that the beam eventually passes through the subsequent system as parallel light. It also disperses the beam energy to avoid excessive power from damaging other components and ensures that the beam irradiated on the reflective amplitude-type SLM can fill the entire amplitude modulation array.

[0127] HWP is used to change the polarization direction of the linear polarized light output by the laser, and modulate the polarization state of the light beam incident on the amplitude modulation array into p-polarized light according to the use requirements of the reflective amplitude-type SLM, so as to ensure that the reflective amplitude-type SLM can work normally;

[0128] PBS is used to split the incident light into two perpendicular linearly polarized lights; the p-polarized light passes completely, while the s-polarized light is reflected at an angle of 45°, and the outgoing direction is 90° to the p-polarized light. Both the p-polarized light and the s-polarized light are linearly polarized lights, and their polarization directions are perpendicular to each other.

[0129] Reflective amplitude SLM is used to modulate the amplitude of incident p-polarized light. It consists of a sandwich structure consisting of a photoconductive layer, a dielectric reflector, a liquid crystal layer, and a transparent conductive electrode (ITO) on a glass substrate. Many basic independent units form a two-dimensional array liquid crystal panel. The processing module can control the voltage to adjust the rotation direction of the liquid crystal molecules through software to achieve amplitude control of a single pixel. The reflected light of the modulated array is s-polarized light.

[0130] Target acquisition detector, used to obtain target images in application scenarios in real time for subsequent system processing;

[0131] A processing system, used to transmit the target image acquired by the target acquisition detector to the reflective amplitude-type SLM, and to transmit the preset phase modulation parameters to the transmissive phase-type SLM, and to simultaneously acquire the classification recognition signal acquired by the classification recognition detector;

[0132] The Fourier lens is used to convert the optical signal in the spatial domain into the optical signal in the spatial frequency domain, so that the subsequent modulation of the optical signal is performed in the spatial frequency domain. It has the same focal length f as the inverse Fourier lens, forming a 4f system;

[0133] The first DONN is used to filter, process and modulate the optical signal in the spatial frequency domain. Its specific parameters are obtained through computer simulation training based on the data set and are manufactured by 3D printing or lithography.

[0134] A transmissive phase-type SLM for performing a nonlinear activation operation on the spatial frequency domain optical signal after the first DONN filtering;

[0135] The anti-Fourier lens is used to convert the optical signal in the spatial frequency domain into the optical signal in the spatial domain, so that the subsequent processing and identification of the optical signal is performed in the spatial domain. It has the same focal length f as the Fourier lens, forming a 4f system;

[0136] The second DONN is used to identify the filtered spatial domain optical signal. Its specific parameters are obtained through computer simulation training based on the data set and are manufactured by 3D printing or lithography.

[0137] The classification and recognition detector is used to collect the optical signal after the second DONN classification and recognition, and transmit the optical signal to the processing system for recognition and processing.

[0138] The present application also proposes a readable storage medium storing a computer program. When the computer program is executed by a processor, the processor executes the following steps:

[0139] Obtain continuous and stable monochromatic p-polarized light;

[0140] Acquire a target image and load the target image into an amplitude modulation array, modulate the monochromatic p-polarized light using the amplitude modulation array loaded with the target image, and obtain a spatial and temporal domain optical signal of the target image containing a complex background;

[0141] The spatial time domain optical signal of the target image containing a complex background is converted into a spatial frequency domain signal, wherein the spatial frequency domain signal is a spatial frequency distribution diagram containing the spatial information of the target image;

[0142] Removing background information of the spatial frequency domain signal to obtain the spatial frequency domain signal after the background is removed;

[0143] The spatial frequency domain signal after background removal is converted into a spatial time domain optical signal after background removal, wherein the spatial time domain optical signal after background removal is a spatial intensity distribution diagram containing spatial information of the target graphic;

[0144] Classify the target image corresponding to the spatial and temporal domain light signal after removing the background.

[0145] The present application also proposes a computer device, including a memory and a processor, wherein the memory stores a computer program, and the computer program is executed by the processor to perform the following steps:

[0146] Obtain continuous and stable monochromatic p-polarized light;

[0147] Acquire a target image and load the target image into an amplitude modulation array, modulate the monochromatic p-polarized light using the amplitude modulation array loaded with the target image, and obtain a spatial and temporal domain optical signal of the target image containing a complex background;

[0148] The spatial time domain optical signal of the target image containing a complex background is converted into a spatial frequency domain signal, wherein the spatial frequency domain signal is a spatial frequency distribution diagram containing the spatial information of the target image;

[0149] Removing background information of the spatial frequency domain signal to obtain the spatial frequency domain signal after the background is removed;

[0150] The spatial frequency domain signal after background removal is converted into a spatial time domain optical signal after background removal, wherein the spatial time domain optical signal after background removal is a spatial intensity distribution diagram containing spatial information of the target graphic;

[0151] Classify the target image corresponding to the spatial and temporal domain light signal after removing the background.

[0152] Those skilled in the art will understand that all or part of the processes in the above-mentioned embodiments can be completed by instructing the relevant hardware through a computer program, and the program can be stored in a non-volatile computer-readable storage medium. When the program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided in the present application may include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. As an illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM).

[0153] The technical features of the above embodiments can be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0154] The above-mentioned embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the patent of the present application. It should be pointed out that, for ordinary technicians in this field, several modifications and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. The above disclosure is only a preferred embodiment of the present invention, and of course it cannot be used to limit the scope of the rights of the present invention. Therefore, equivalent changes made according to the claims of the present invention are still within the scope covered by the present invention.

Claims

1. A target image classification method based on complex background, characterized in that: The method comprises: Obtain continuous and stable monochromatic p-polarized light; Acquire a target image, and load the target image into an amplitude modulation array, and modulate the monochromatic p-polarized light using the amplitude modulation array after the target image is loaded, so as to obtain a spatial time-domain optical signal of the target image containing a complex background; Converting the spatial time domain optical signal of the target image containing the complex background into a spatial frequency domain signal, wherein the spatial frequency domain signal is a spatial frequency distribution diagram containing the spatial information of the target image; Removing background information of the spatial frequency domain signal to obtain a spatial frequency domain signal after the background is removed; Converting the background-removed spatial frequency domain signal into a background-removed spatial time domain optical signal, wherein the background-removed spatial time domain optical signal is a spatial intensity distribution diagram containing spatial information of the target graphic; The target image corresponding to the spatial-temporal optical signal after the background is removed is classified.

2. The target image classification method based on complex background according to claim 1 is characterized in that: The step of obtaining continuous and stable monochromatic p-polarized light specifically includes: The collection laser emits a continuous and stable monochromatic linearly polarized laser beam; Passing the monochromatic linearly polarized laser beam through a beam expanding lens pair and adjusting the lens pair to obtain a parallel beam after beam expansion; Performing polarization modulation on the expanded parallel light beam according to a half-wave plate; The angle of the half-wave plate is controlled to obtain continuous and stable monochromatic p-polarized light.

3. The target image classification method based on complex background according to claim 2 is characterized in that: The step of acquiring a target image and loading the target image into an amplitude modulation array further comprises: performing polarization filtering on the continuous and stable monochromatic p-polarized light through a polarization beam splitter prism PBS.

4. The target image classification method based on complex background according to claim 3 is characterized in that: The method of using the amplitude modulation array after loading the target image to modulate the monochromatic p-polarized light to obtain the spatial-temporal light signal of the target image containing a complex background also includes: distributing the light field intensity of the reflected light according to the grayscale value of the loaded grayscale image to obtain the spatial-temporal light signal of the target image containing a complex background.

5. The target image classification method based on complex background according to claim 4 is characterized in that: The step of converting the spatial time domain optical signal of the target image containing a complex background into a spatial frequency domain signal specifically includes: The amplitude modulation array of the reflective amplitude-type spatial light modulator SLM receives the monochromatic p-polarized light reflected by the first polarization beam splitter prism PBS, and the monochromatic p-polarized light is reflected as monochromatic s-polarized light after amplitude modulation; The first polarization beam splitter prism PBS reflects the monochromatic s-polarized light in a direction perpendicular to the incident light path; The reflected light beam passes through the first polarization beam splitter prism PBS from the amplitude modulation array, and the focal length f of the lens is determined according to the distance of the light beam in its propagation path when it reaches the Fourier lens; When the spatial time domain optical signal of the target image with a complex background passes through the Fourier lens, the Fourier lens converts it into the spatial frequency domain to obtain a spatial frequency domain signal.

6. The target image classification method based on complex background according to claim 5 is characterized in that: The step of converting the spatial time domain optical signal of the target image containing a complex background into a spatial frequency domain signal specifically includes: Perform a two-dimensional Fourier transform on the modulated s-polarized light to convert the spatial information of the target image into frequency information; According to the spatial frequency distribution diagram formed on the focal plane, the spatial frequency domain signal of the target image containing a complex background is determined by the signal after Fourier transformation.

7. The target image classification method based on complex background according to claim 6 is characterized in that: The removing background information of the spatial frequency domain signal to obtain the spatial frequency domain signal after the background is removed specifically includes: Performing filtering processing on the spatial frequency domain signal according to a first diffractive optical neural network DONN; The spatial frequency domain signal after filtering is nonlinearly activated according to the transmissive phase-type spatial light modulator SLM to obtain the processed spatial frequency domain signal.

8. A target image classification system based on complex background, characterized in that: The system comprises: a polarization acquisition unit, a time domain signal acquisition unit, a signal processing unit and a classification unit; The polarization acquisition unit is used to acquire continuous and stable monochromatic p polarization; The time domain signal acquisition unit is used to collect a target image, load the target image into an amplitude modulation array, and modulate the monochromatic p-polarized light using the amplitude modulation array loaded with the target image to acquire a spatial time domain optical signal of the target image containing a complex background; The signal processing unit is used to convert the spatial time domain light signal of the target image containing a complex background into a spatial frequency domain signal, wherein the spatial frequency domain signal is a spatial frequency distribution diagram containing spatial information of the target image; remove the background information of the spatial frequency domain signal to obtain the spatial frequency domain signal after the background is removed; convert the spatial frequency domain signal after the background is removed into a spatial time domain light signal after the background is removed, wherein the spatial time domain light signal after the background is removed is a spatial intensity distribution diagram containing spatial information of the target graphic; and the classification unit is used to classify the target image corresponding to the spatial time domain light signal after the background is removed.

9. A readable storage medium storing a computer program, wherein when the computer program is executed by a processor, the processor is caused to execute the steps of the method according to any one of claims 1 to 7.

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 executes the steps of the method according to any one of claims 1 to 7.

Images