Multi-dimensional detection verification converter for planar optical super-diffraction imaging research
By introducing metasurface optical components and multi-dimensional detection modules into planar optical imaging technology, combining adversarial networks and feature extraction algorithms, the problem of poor imaging quality in traditional technologies is solved, and high-resolution imaging and accuracy of super-diffraction limits is achieved promote.
Patent Information
- Application Number
- CN202411243801.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-09-05
- Publication Date
- 2024-12-13
- Estimated Expiration
- Not applicable · inactive patent
AI Technical Summary
Traditional planar imaging technology is difficult to achieve high-resolution imaging with super-diffraction limits, unable to effectively obtain multi-dimensional information of the beam, and lacks image processing algorithms and verification methods, resulting in poor imaging quality.
Design a multi-dimensional detection verification converter, including a metasurface optical element, a multi-dimensional detection module and a data processing unit, to realize the phase and amplitude modulation of the beam through the metasurface optical element, the multi-dimensional detection module captures the beam information and uses an adversarial network and feature extraction. The algorithm improves imaging quality.
It realizes multi-dimensional information acquisition of the beam in spatial position, polarization state and wavefront state, breaks through the diffraction limit of traditional imaging technology, improves the accuracy and reliability of imaging, and reduces research costs.
Smart Images
Figure CN119124355A_ABST
Abstract
Description
Technical Field
[0001] The invention relates to the technical field of optical imaging, and in particular to a multi-dimensional detection and verification converter used for planar optical super-diffraction imaging research. Background Art
[0002] Traditional planar imaging technology is a technology widely used in the field of optical imaging, including microscopes, telescopes, etc.; it mainly relies on optical elements such as lenses to collect and focus light to form images.
[0003] Traditional imaging technology is restricted by the diffraction limit and it is difficult to achieve high-resolution imaging beyond the diffraction limit. The imaging system in the existing technology uses a single detector and cannot obtain multi-dimensional information on the spatial position, polarization state, and wavefront state of the light beam; it lacks effective image processing algorithms and cannot restore high-resolution images from low-resolution images, and cannot guarantee imaging quality; it lacks effective image quality verification methods and cannot evaluate the accuracy and reliability of the reconstructed image. Summary of the invention
[0004] 1. Technical issues to be resolved
[0005] In view of the deficiencies in the prior art, the present invention provides a multi-dimensional detection and verification converter for planar optical super-diffraction imaging research, which solves the problems of how to obtain light beam information through multiple dimensions and how to achieve imaging accuracy and reliability through algorithms.
[0006] (II) Technical solution
[0007] To achieve the above objectives, the present invention is implemented through the following technical solutions: a multi-dimensional detection verification converter for planar optical super-diffraction imaging research, comprising:
[0008] At least one metasurface optical element for phase modulation and amplitude modulation of an incident light beam to achieve imaging beyond the diffraction limit and to surpass the diffraction limit in a conventional optical imaging system to achieve higher resolution;
[0009] A multi-dimensional detection module, which is used to capture the light beam information modulated by the metasurface optical element and convert it into an electrical signal, including multi-dimensional beam splitting technology, polarization detector and wavefront detector;
[0010] a data processing unit for receiving the electrical signals from the multi-dimensional detection module, and processing and verifying them, combining adversarial networks and feature extraction to improve the imaging quality;
[0011] A user interface for a user to interact with a multi-dimensional detection verification converter for planar optical super-diffraction imaging.
[0012] Preferably, a metasurface optical element is characterized in that the size of the nanostructure unit in the metasurface optical element is 1 / 50 to 1 / 100 of the wavelength of the working light to achieve an imaging effect beyond the diffraction limit, wherein the wavelength of the working light is in the visible light band; the material of the metasurface optical element is a transparent material with a high refractive index, such as titanium dioxide, to optimize the electromagnetic response; the design of the metasurface optical element is based on a physical model B=μH, wherein B represents the magnetic induction intensity, H represents the magnetic field intensity, and μ represents the magnetic permeability; the magnetic induction intensity of the metasurface optical element is determined by the physical model, and the geometric parameters and sub-wavelength structure of the metasurface optical element are determined by simulation software, so that the response of each nanostructure unit to the electromagnetic wave meets the design requirements; electromagnetic simulation is performed using simulation software to further determine the metasurface light The geometric parameters of the optical element, including the unit cell size less than the sub-wavelength, shape, and periodic two-dimensional array, are used to manufacture the metasurface optical element through the simulation results, and the optical test platform is used to test the beam diffraction efficiency. The test results verify whether the sub-wavelength structure of the metasurface optical element can meet the requirements of super-diffraction limit imaging; the metasurface optical element is integrated into the multi-dimensional detection verification converter of planar optical super-diffraction imaging to achieve high-resolution imaging; if the product of the aperture of the optical system of a traditional system and the refractive index of the medium is 1.0, after using the metasurface optical element, the design and manufacturing accuracy of the metasurface optical element is experimentally simulated according to the simulation software, and the product of the aperture of the optical system and the refractive index of the medium is increased to 1.5. If the working light wavelength is 500nm, the traditional resolution is The resolution after using metasurface optics is After using metasurface optical elements, the resolution was improved from 622 nanometers to 415 nanometers. After using metasurface optical elements, the resolution was improved from 622 nanometers to 415 nanometers, which significantly improved the resolution of imaging.
[0013] Preferably, the multi-dimensional detection module can capture and process light beam information at different spatial positions and polarization directions through multi-dimensional beam splitting technology, and the multi-dimensional beam splitting technology includes a multi-detector array, a polarization beam splitter, a wavefront encoder and data acquisition and processing; the multi-detector array is distributed at different spatial positions of the metasurface optical element to capture information about the light beam at the spatial position; the polarization beam splitter is integrated in the multi-dimensional detection module to split the polarization state of the light beam, split the incident light beam into sub-beams with multiple polarization states, and guide the sub-beams split into multiple polarization states to the polarization detector, so as to obtain information about the polarization of the light beam; the wavefront encoder is integrated in the multi-dimensional detection module to encode the wavefront of the light beam, split the incident light beam into multiple wavefront states The sub-beams in the state are divided into multiple wavefront states and guided to the wavefront detector. The wavefront detector reconstructs the wavefront state of the entire beam by measuring the wavefront states of multiple sub-beams, thereby realizing high-precision super-diffraction imaging. The multi-dimensional detection module collects data from the polarization detector and the wavefront detector, converts the sub-beam data into electrical signals through the internal photoelectric conversion elements, records the electrical signals of the multi-dimensional detection module in real time, and processes the collected data to realize comprehensive capture and analysis of the beam information, and adjusts the beam according to the beam information to improve its imaging quality, wherein the processing of the collected data includes extracting the spatial position, polarization state, and wavefront state of the beam, wherein the polarization state is the directional vibration of the light wave, and the wavefront state is the phase distribution of the light wave in space.
[0014] Preferably, the data processing unit includes: an image reconstruction algorithm that uses an adversarial network to restore a high-resolution image from a low-resolution image, and adds a sharpening filter as a connection between the generator and the discriminator in the adversarial network, processes the image output of the generator, and uses the processed image as the input of the discriminator to make the image reconstruction algorithm more accurate; a verification algorithm that uses feature extraction to verify the quality and accuracy of the reconstructed image.
[0015] Preferably, the image reconstruction algorithm is based on an adversarial network, and a large number of low-resolution and high-resolution image pairs are collected as training data, so that the image pairs have diversity in content, style and resolution, which is used to improve the generalization ability of the model; an adversarial network model is designed, including a generator, a sharpening filter and a discriminator; the generator generates a high-resolution image from a low-resolution image, and the high-resolution image is input into the sharpening filter, and the sharpening filter is processed by a Laplacian operator. Enhance the edges and clarity of the image generated by the generator, make the image generated by the generator clearer and sharper, improve the quality of the image generated by the generator, and make it more realistic; the discriminator is responsible for judging whether the image input from the sharpening filter is real or generated, where I(x,y) represents the brightness of the high-resolution image at point (x,y), is the local curvature of brightness changes in high-resolution images; the model is trained using low-resolution images as input and high-resolution images as targets; the discriminator extracts the features of the image after optimization by the sharpening filter through the internal convolutional layer, and outputs a judgment value based on the extracted features, indicating the probability that the input image is real; the high-resolution image after optimization by the sharpening filter is more difficult for the discriminator to distinguish, forcing the generator to generate higher quality images, thereby improving the performance of the entire adversarial network; the trained model is applied to the light beam information modulated by the multi-dimensional detection module, and the quality of the reconstructed image is improved by restoring the low resolution to the high-resolution image.
[0016] Preferably, a verification method constructs a scale space by using the ORB algorithm in feature extraction and uses different kernel functions Obtain a blurred version of the image in scale space, detect key points, i.e., local extreme points, in scale space, where x and y represent the position coordinates in the image, σ is the standard deviation of the kernel, which controls the degree of blur, and θ is the angle parameter; calculate the direction of each key point, determine the position of the key point in scale space, and thus identify the corners, edges, and shape features of the reconstructed image generated in the adversarial network and the original high-resolution image; the ORB algorithm calculates the feature matching between the reconstructed image generated by the adversarial network and the original high-resolution image through a brute force matching method, compares the features in the two images, and finds the best match; by Calculate the Euclidean distance between features. The smaller the distance, the higher the feature similarity. d(F1, F2) is the Euclidean distance between two features F1 and F2. 1i and F 2i are the values of features F1 and F2 in the i-th dimension respectively; n is the number of dimensions of the feature; the image quality is evaluated by analyzing the similarity. If the feature similarity between the reconstructed image and the original image is very high, the image quality output by the generator is close to the original image, with higher resolution and detail fidelity; if the similarity is low, the generated image has quality problems, and it is necessary to further optimize the network structure and training process of the image reconstruction algorithm, and display the high-resolution image to the user through the user interface.
[0017] Preferably, a user interface is provided for interacting with a user, receiving input instructions from the user, and displaying the processed image, and operating the quality of the processed image through a window on the user interface, such as zooming, panning, or comparing the original image with the processed image.
[0018] (III) Beneficial effects
[0019] The invention provides a multi-dimensional detection verification converter for planar optical super-diffraction imaging research.
[0020] It has the following beneficial effects:
[0021] 1. The present invention can simultaneously obtain information on the spatial position, polarization state, wavefront state and other dimensions of the light beam. The acquisition of multi-dimensional information helps to more comprehensively understand the characteristics of the light beam and provide richer data for subsequent image processing and analysis. Combined with an effective image processing algorithm, it can break through the diffraction limit of traditional imaging technology and achieve high-resolution imaging beyond the diffraction limit.
[0022] 2. The present invention reduces the number of times experiments or tests need to be repeated due to poor imaging quality by improving the accuracy and reliability of imaging, thereby reducing the cost of research and production; in addition, the present invention can be applied not only to traditional planar optical imaging fields, such as microscopes and telescopes, but also to other fields requiring high-resolution imaging, such as biomedical imaging, material science, precision engineering, etc. BRIEF DESCRIPTION OF THE DRAWINGS
[0023] Figure 1 It is a schematic diagram of the process of the present invention; DETAILED DESCRIPTION
[0024] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.
[0025] The core component of the multi-dimensional detection verification converter for planar optical super-diffraction imaging research is a metasurface optical element, which is made of a transparent material with a high refractive index, such as titanium dioxide, and has a subwavelength structure, that is, its size is much smaller than the wavelength of the working light, usually in the range of 1 / 50 to 1 / 100 of the wavelength of the working light; the precise design of these nanostructure units is based on the physical model B=μH, and the electromagnetic simulation software is used to simulate and use the physical model to determine the magnetic induction intensity, geometric parameters and subwavelength structure of each unit, including unit cell size, shape and period; the manufacture and testing of the metasurface optical element ensures that it can accurately phase modulate and amplitude modulate the incident light beam, thereby achieving imaging effects beyond the diffraction limit.
[0026] The metasurface optical element is combined with a multi-dimensional detection module, which consists of a multi-detector array, a polarization beam splitter, a wavefront encoder and a data acquisition and processing system; the multi-detector array is distributed at different spatial positions of the metasurface optical element to capture the spatial position information of the light beam, while the polarization beam splitter and the wavefront encoder are responsible for the segmentation and encoding of the polarization state and wavefront state of the light beam, respectively, and guide sub-beams of different states to the corresponding polarization detectors and wavefront detectors; this module can record the output electrical signal in real time and perform data processing to comprehensively capture and analyze the light beam information.
[0027] The data processing unit is another key part of the multidimensional detection and verification converter in planar optical super-diffraction imaging research. It obtains the electrical signal after processing the light beam information from the multidimensional detection and verification converter and performs processing and analysis. It includes an image reconstruction algorithm and a verification algorithm; the image reconstruction algorithm is based on an adversarial network and is trained using a large number of image pairs from low resolution to high resolution to improve the generalization ability of the model; the algorithm includes a generator, a sharpening filter and a discriminator. The generator is responsible for generating high-resolution images from low-resolution images, and the sharpening filter enhances the edges and clarity of the image through the Laplace operator, while the discriminator is used to distinguish between the image processed by the sharpening filter and the real image. This process improves the quality of image reconstruction through continuous training and optimization.
[0028] The verification algorithm uses feature extraction technology, especially the ORB algorithm, to evaluate the quality of the reconstructed image; the ORB algorithm constructs a scale space and detects key points, calculates the direction and position of key points, identifies corners, edges and shape features in the image, and The Euclidean distance between features is calculated. The smaller the distance, the higher the feature similarity. The similarity and quality of the images are evaluated by comparing the feature matching and Euclidean distance between the reconstructed image and the original high-resolution image.
[0029] The user interface provides users with an intuitive operating platform for inputting commands, displaying processed images, and providing feedback and control; users can use the interface to operate high-resolution images processed by the verification algorithm, such as zooming and panning, and compare the original image with the processed image to intuitively evaluate the imaging results.
[0030] In practical applications, the planar optical super-diffraction imaging method of the multi-dimensional detection verification converter in the planar optical super-diffraction imaging research includes the following steps: first, the incident light beam is modulated using a metasurface optical element, and then the modulated light beam information is captured through a multi-dimensional detection module, and then the captured information is converted into an electrical signal and input into a data processing unit for analysis and processing, and finally the imaging result is output through a user interface.
[0031] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.
Claims
1. A multi-dimensional detection verification converter for planar optical super-diffraction imaging research, characterized in that: include: At least one metasurface optical element for phase modulation and amplitude modulation of an incident light beam to achieve imaging beyond the diffraction limit and to surpass the diffraction limit in a conventional optical imaging system to achieve higher resolution; A multi-dimensional detection module, which is used to capture the light beam information modulated by the metasurface optical element and convert it into an electrical signal, including multi-dimensional beam splitting technology, polarization detector and wavefront detector; a data processing unit for receiving the electrical signals from the multi-dimensional detection module, and processing and verifying them, combining adversarial networks and feature extraction to improve the imaging quality; A user interface for a user to interact with a multi-dimensional detection verification converter for planar optical super-diffraction imaging.
2. The multi-dimensional detection verification converter for planar optical super-diffraction imaging research according to claim 1, characterized in that: The metasurface optical element determines the electromagnetic response of the metasurface element through imaging requirements, establishes a physical model B=μH, and obtains the magnetic properties of the metasurface optical element material and the degree of influence on the electromagnetic response according to the physical model, wherein B is the magnetic induction intensity, H is the magnetic field intensity, and μ is the magnetic permeability; uses simulation software to design the subwavelength structure of the metasurface element, that is, a structure in which the element size is smaller than the wavelength of light, and controls the light wave through the magnetic induction intensity obtained by the physical model and the subwavelength structure designed by the simulation software; after simulation, manufactures the metasurface element for performance testing to verify whether the subwavelength of the metasurface optical element meets the requirements of super-diffraction limit imaging; and integrates the metasurface element into a multi-dimensional detection verification converter for planar optical super-diffraction imaging.
3. The multi-dimensional detection verification converter for planar optical super-diffraction imaging research according to claim 1, characterized in that: The multi-dimensional detection module can capture beam information at different spatial positions and polarization directions through multi-dimensional beam splitting technology, which includes a multi-detector array, a polarization beam splitter, a wavefront encoder, and data acquisition and processing; The multi-detector array is distributed at the edge and center of the metasurface optical element to capture the information of the spatial position of the light beam; the polarization beam splitter is integrated in the multi-dimensional detection module to split the polarization state of the light beam, split the incident light beam into sub-beams of multiple polarization states, and guide the sub-beams split into multiple polarization states to the polarization detector, so as to obtain information about the polarization of the light beam; The wavefront encoder is integrated in the multi-dimensional detection module, and is used to encode the wavefront of the light beam, split the incident light beam into sub-beams with multiple wavefront states, and guide the sub-beams split into multiple wavefront states to the wavefront detector. The wavefront detector reconstructs the wavefront state of the entire light beam by measuring the wavefront states of multiple sub-beams, thereby achieving high-precision super-diffraction imaging; the multi-dimensional detection module collects data from the polarization detector and the wavefront detector, and converts the sub-beam data into electrical signals through the internal photoelectric conversion elements, records the electrical signals of the multi-dimensional detection module in real time, and processes the collected data to achieve comprehensive capture and analysis of the light beam information, and adjust the light beam according to the light beam information to improve its imaging quality, wherein the processing of the collected data includes extracting the spatial position, polarization state, and wavefront state of the light beam, wherein the polarization state is the directional vibration of the light wave, and the wavefront state is the phase distribution of the light wave in space.
4. The multi-dimensional detection verification converter for planar optical super-diffraction imaging research according to claim 1, characterized in that: The data processing unit comprises: An image reconstruction algorithm that uses adversarial networks to recover high-resolution images from low-resolution images; A verification algorithm that uses feature extraction to verify the quality and accuracy of the reconstructed image.
5. The multi-dimensional detection and verification converter for planar optical super-diffraction imaging research according to claim 4, characterized in that: The image reconstruction algorithm is based on an adversarial network, and a large number of low-resolution and high-resolution image pairs are collected as training data, so that the image pairs have diversity in content, style and resolution, which is used to improve the generalization ability of the model; an adversarial network model is designed, including a generator, a sharpening filter and a discriminator; The generator generates a high-resolution image from a low-resolution image, and inputs the high-resolution image into a sharpening filter. The sharpening filter is processed by the Laplacian operator L(x,y) = Enhance the edges and clarity of the high-resolution image generated by the generator, make the image generated by the generator clearer and sharper, improve the quality of the image generated by the generator, and make it more realistic, where I(x,y) represents the brightness of the high-resolution image at point (x,y), is the local curvature of brightness changes in high-resolution images; the discriminator is responsible for judging whether the image input from the sharpening filter is real or generated; the model is trained using low-resolution images as input and high-resolution images as targets; the discriminator distinguishes high-resolution images optimized by the sharpening filter, thereby improving the performance of the entire adversarial network; the trained model is applied to the light beam information modulated by the multi-dimensional detection module, the low-resolution image is restored to the high-resolution image, and the quality of the reconstructed image is improved.
6. The multi-dimensional detection verification converter for planar optical super-diffraction imaging research according to claim 5, characterized in that: The verification method constructs a scale space by using the ORB algorithm in feature extraction and uses different kernel functions. Obtain a blurred version of the image in scale space and detect key points, i.e. local extreme points, in scale space, where x and y represent the position coordinates in the image, σ is the standard deviation of the kernel, which controls the degree of blur, and θ is the angle parameter used to rotate the kernel to adapt to the local directionality of the image; Calculate the direction of each key point and determine the position of the key point in the scale space, so as to identify the corners, edges, and shape features of the reconstructed image and the original high-resolution image generated by the adversarial network; The ORB algorithm calculates the Euclidean distance between features through brute force matching to determine the matching similarity, analyzes the distribution of similarity to evaluate the image quality, and displays the verified image to the user through the user interface.
7. The multi-dimensional detection verification converter for planar optical super-diffraction imaging research according to claim 1, characterized in that: The user interface is used to interact with the user, receive the user's input instructions, display the processed images, and provide intuitive feedback and control.