Non-visible light wave band high-speed imaging method and system based on metasurface lens and medium

Abstract

The invention provides an invisible light wave band high-speed imaging method and system based on a metasurface lens and a medium, and the method comprises the steps: outputting terahertz or ultraviolet wave band continuous pulse illumination based on an invisible light source, and obtaining an illumination light source; regulating and controlling the irradiation light source based on the metasurface lens to obtain an incident light source; irradiating an incident light source to the surface of a target, and collecting a target reflection image sequence based on a high-speed focal plane array detector; preprocessing the target reflection image sequence based on a sparse image processing algorithm to obtain a target reflection image of a complete frame sequence; reconstructing the target reflection image of the complete frame sequence to obtain a reconstructed image, and transmitting the reconstructed image to a terminal according to a preset mode for display; after focusing regulation and control are carried out through the metasurface lens array, the light irradiates the target surface and is collected by the high-speed focal plane detector, the response time is nanosecond, the integrated miniature metasurface lens array with the high-speed process is adapted, and the wavefront regulation and control and focusing efficiency is improved.

Description

Technical Field

[0001] This application relates to the field of imaging technology, and more specifically, to a high-speed imaging method, system, and medium for the non-visible light band based on metasurface lenses. Background Technology

[0002] Non-visible light (such as the THz band: 0.1–10 THz, ultraviolet (UV): 200–400 nm) imaging has the following unique advantages:

[0003] It can penetrate visible light-opaque materials such as silicon, ceramics, and biological tissues; it matches the resonance frequency of material molecules and can identify chemical bonding states; the high-energy short wavelength of ultraviolet light is beneficial to improving imaging resolution; THz imaging is safe, non-ionizing, and suitable for biological detection.

[0004] However, non-visible light imaging still faces the following technical challenges: slow response of imaging devices and low frame rate (<1kfps); inability to achieve unified control across a wide frequency band (UV+THz); large size of imaging systems, complex optical components, and poor wavefront control; and a lack of structural innovations to meet the needs of high-speed process detection and biological dynamic imaging. Summary of the Invention

[0005] The purpose of this application is to provide a high-speed imaging method, system, and medium for the non-visible light band based on metasurface lenses. After focusing and controlling the image through a metasurface lens array, the image is irradiated onto the target surface and acquired by a high-speed focal plane detector. The response time is in the nanosecond range, which is suitable for high-speed processes. The integrated micro metasurface lens array improves wavefront control and focusing efficiency.

[0006] This application also provides a high-speed imaging method for the non-visible light band based on metasurface lenses, including:

[0007] An illumination source is obtained by outputting continuous pulse illumination in the terahertz or ultraviolet band from a non-visible light source;

[0008] An incident light source is obtained by controlling the illumination light source using metasurface lenses;

[0009] An incident light source is irradiated onto the target surface, and a sequence of target reflection images is acquired based on a high-speed focal plane array detector.

[0010] The target reflection image sequence is preprocessed based on the sparse image processing algorithm to obtain the target reflection image of the complete frame sequence;

[0011] The target reflection image of the complete frame sequence is reconstructed to obtain the reconstructed image, which is then transmitted to the terminal for display according to a preset method.

[0012] Optionally, in the high-speed imaging method for the non-visible light band based on metasurface lenses described in this application embodiment, the illumination source is obtained by outputting terahertz or ultraviolet band continuous pulse illumination from a non-visible light source, specifically including:

[0013] The illumination source is obtained by outputting continuous pulse illumination in the terahertz or ultraviolet band from a non-visible light source, specifically including:

[0014] Based on the continuous wave output in the terahertz band by quantum cascade laser, the continuous wave is converted into a pulse sequence based on electro-optic modulation;

[0015] Terahertz pulsed illumination can be obtained by adjusting the pulse width based on the pulse sequence;

[0016] Based on the ultraviolet laser emitting ultraviolet light, the pulse parameters of the ultraviolet light are set, including the pulse frequency and pulse band;

[0017] Continuous pulse illumination in the ultraviolet light band is obtained based on the pulse frequency and pulse band of ultraviolet light;

[0018] The illumination source is obtained by weighted fusion of terahertz pulsed illumination and ultraviolet continuous pulsed illumination.

[0019] Optionally, in the high-speed imaging method for the non-visible light band based on metasurface lenses described in the embodiments of this application, the metasurface broadband lens design method is as follows:

[0020] Metasurface lenses achieve holographic manipulation of wavefronts by arranging nanostructures on a subwavelength scale. The phase response is analyzed using the principle of equal phase delay, as shown in the following formula:

[0021]

[0022] in:

[0023] φ(x,y,λ): The phase response of the metasurface unit at wavelength λ;

[0024] (x0, y0): Coordinates of the focus;

[0025] f: focal length;

[0026] λ: Incident wavelength (200nm–3mm);

[0027] The metasurface material employs a two-dimensional periodic array, and the phase delay is automatically adjusted according to λ to form a lens array with controllable dispersion.

[0028] Optionally, in the high-speed imaging method for non-visible light band based on metasurface lenses described in this application embodiment, the incident light source is irradiated onto the target surface, and the target reflection image sequence is acquired based on a high-speed focal plane array detector, specifically including:

[0029] Based on the incident visible light signal obtained from the incident light source, let the incident non-visible light signal be:

[0030] I in (x,y,λ,t)=I0(x,y,λ)·e -μ(λ)d

[0031] I0: Initial incident irradiance;

[0032] μ(λ): Wavelength-dependent absorption coefficient;

[0033] d: Thickness of the material being measured;

[0034] After being manipulated by a metasurface lens, the image is focused onto the plane of a high-speed focal plane array detector, resulting in a sequence of target reflection images, as shown in the following formula:

[0035]

[0036] T(λ): Transmittance of the metasurface;

[0037] δ(φ): Phase matching function;

[0038] I foc : Brightness of the imaging point.

[0039] Optionally, in the high-speed imaging method for non-visible light bands based on metasurface lenses described in this application embodiment, the target reflection image sequence is preprocessed based on a sparse image processing algorithm to obtain a complete frame sequence of target reflection images, specifically including:

[0040] The target reflection image sequence is obtained, and the target reflection image sequence is decomposed into low-frequency approximate components and high-frequency detail components based on wavelet transform.

[0041] Based on the high-frequency detail components, sparse distribution information is obtained. Based on the Fourier transform, the low-frequency approximation classification of the low-frequency region is transformed into the frequency domain to obtain the sparse coefficients of the high-frequency region.

[0042] By quantizing and encoding the sparse distribution information and the sparse coefficients in the high-frequency region, compressed sparsity is obtained.

[0043] Based on compression sparse analysis, compression ratio and distortion state information are balanced and distributed to obtain the optimal compression ratio, and target reflection images of complete frame sequences are generated.

[0044] Optionally, in the high-speed imaging method for the non-visible light band based on metasurface lenses described in this application embodiment, the target reflection image of the complete frame sequence is reconstructed to obtain a reconstructed image, and the reconstructed image is transmitted to the terminal for display in a preset manner, specifically including:

[0045] Image decompression of target reflection images of complete frame sequences is performed based on sparse regularization constraints to obtain decompressed data;

[0046] Set a segmentation threshold, and perform noise analysis on the pressurized data based on the segmentation threshold to obtain noise data;

[0047] Noise is removed from the data to obtain noise-free data. Dynamic reconstruction is then performed based on the noise-free data to obtain the reconstructed image.

[0048] Secondly, embodiments of this application provide a high-speed imaging system for the non-visible light band based on metasurface lenses. This system includes a memory and a processor. The memory includes a program for a high-speed imaging method for the non-visible light band based on metasurface lenses. When the program for the high-speed imaging method for the non-visible light band based on metasurface lenses is executed by the processor, it implements the following steps:

[0049] An illumination source is obtained by outputting continuous pulse illumination in the terahertz or ultraviolet band from a non-visible light source;

[0050] An incident light source is obtained by controlling the illumination light source using metasurface lenses;

[0051] An incident light source is irradiated onto the target surface, and a sequence of target reflection images is acquired based on a high-speed focal plane array detector.

[0052] The target reflection image sequence is preprocessed based on the sparse image processing algorithm to obtain the target reflection image of the complete frame sequence;

[0053] The target reflection image of the complete frame sequence is reconstructed to obtain the reconstructed image, which is then transmitted to the terminal for display according to a preset method.

[0054] Optionally, in the high-speed imaging system for the non-visible light band based on metasurface lenses described in this application embodiment, the illumination source is obtained by outputting terahertz or ultraviolet band continuous pulse illumination from a non-visible light source, specifically including:

[0055] The illumination source is obtained by outputting continuous pulse illumination in the terahertz or ultraviolet band from a non-visible light source, specifically including:

[0056] Based on the continuous wave output in the terahertz band by quantum cascade laser, the continuous wave is converted into a pulse sequence based on electro-optic modulation;

[0057] Terahertz pulsed illumination can be obtained by adjusting the pulse width based on the pulse sequence;

[0058] Based on the ultraviolet laser emitting ultraviolet light, the pulse parameters of the ultraviolet light are set, including the pulse frequency and pulse band;

[0059] Continuous pulse illumination in the ultraviolet light band is obtained based on the pulse frequency and pulse band of ultraviolet light;

[0060] The illumination source is obtained by weighted fusion of terahertz pulsed illumination and ultraviolet continuous pulsed illumination.

[0061] Optionally, in the non-visible light band high-speed imaging system based on metasurface lenses described in this application embodiment, the metasurface broadband lens design method is as follows:

[0062] Metasurface lenses achieve holographic manipulation of wavefronts by arranging nanostructures on a subwavelength scale. The phase response is analyzed using the principle of equal phase delay, as shown in the following formula:

[0063]

[0064] in:

[0065] φ(x,y,λ): The phase response of the metasurface unit at wavelength λ;

[0066] (x0, y0): Coordinates of the focus;

[0067] f: focal length;

[0068] λ: Incident wavelength (200nm–3mm);

[0069] The metasurface material employs a two-dimensional periodic array, and the phase delay is automatically adjusted according to λ to form a lens array with controllable dispersion.

[0070] Thirdly, embodiments of this application also provide a computer-readable storage medium, which includes a program for a high-speed imaging method in the non-visible light band based on metasurface lenses. When the program for the high-speed imaging method in the non-visible light band based on metasurface lenses is executed by a processor, it implements the steps of the high-speed imaging method in the non-visible light band based on metasurface lenses as described in any of the preceding claims.

[0071] As can be seen from the above, the non-visible light band high-speed imaging method, system, and medium provided in this application embodiment are as follows: An illumination source is obtained by outputting continuous pulse illumination in the terahertz or ultraviolet band based on a non-visible light source; the illumination source is controlled based on a metasurface lens to obtain an incident light source; the incident light source is irradiated onto the target surface, and a target reflection image sequence is acquired based on a high-speed focal plane array detector; the target reflection image sequence is preprocessed based on a sparse image processing algorithm to obtain a complete frame sequence of target reflection images; the complete frame sequence of target reflection images is reconstructed to obtain a reconstructed image, and the reconstructed image is transmitted to a terminal for display according to a preset method; the target surface is irradiated after focusing and control by a metasurface lens array and acquired by a high-speed focal plane detector, with a response time on the nanosecond level, adaptable to high-speed processes, and an integrated micro metasurface lens array, improving wavefront control and focusing efficiency. Attached Figure Description

[0072] To more clearly illustrate the technical solutions of the embodiments of this application, the accompanying drawings used in the embodiments of this application will be briefly introduced below. It should be understood that the following drawings only show some embodiments of this application and should not be regarded as a limitation of the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0073] Figure 1 A flowchart of a high-speed imaging method for the non-visible light band based on metasurface lenses provided in this application embodiment;

[0074] Figure 2 A diagram of a metasurface lens array structure for a high-speed imaging method in the non-visible light band based on metasurface lenses, provided in an embodiment of this application;

[0075] Figure 3 A schematic diagram of the module structure and signal flow of a high-speed imaging system in the non-visible light band based on metasurface lenses provided in this application embodiment;

[0076] Figure 4 A cross-sectional view of the ultraviolet / terahertz FPA chip structure of a high-speed imaging system based on metasurface lenses in the non-visible light band provided in this application embodiment;

[0077] Figure 5 Comparison of wafer defect imaging effects of a high-speed imaging system based on metasurface lenses in the non-visible light band, provided in the embodiments of this application. Detailed Implementation

[0078] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of the embodiments. The components of the embodiments of this application described and shown in the accompanying drawings can generally be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of this application provided in the accompanying drawings is not intended to limit the scope of the claimed application, but merely represents selected embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without inventive effort are within the scope of protection of this application.

[0079] It should be noted that similar reference numerals and letters in the following figures indicate similar items; therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures. Furthermore, in the description of this application, the terms "first," "second," etc., are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0080] Please refer to Figure 1 , Figure 1 This is a flowchart illustrating a high-speed imaging method for the non-visible light band based on metasurface lenses, as described in some embodiments of this application. This high-speed imaging method for the non-visible light band based on metasurface lenses is used in a terminal device and includes the following steps:

[0081] S101, based on the output of terahertz or ultraviolet band continuous pulse illumination from a non-visible light source, obtains an illumination source;

[0082] S102, the incident light source is obtained by controlling the illumination light source based on the metasurface lens;

[0083] S103, the incident light source is irradiated onto the target surface, and the target reflection image sequence is acquired based on the high-speed focal plane array detector;

[0084] S104, Preprocess the target reflection image sequence based on the sparse image processing algorithm to obtain the target reflection image of the complete frame sequence;

[0085] S105, reconstruct the target reflection image of the complete frame sequence to obtain the reconstructed image, and transmit the reconstructed image to the terminal for display in a preset manner.

[0086] It should be noted that the detector is one of the following:

[0087] Ultraviolet focal plane array, based on AlGaN or ZnO materials, with a response range of 200–400 nm;

[0088] Terahertz focal plane arrays, based on electrothermal detectors or graphene detectors, have a response bandwidth of 0.1–10 THz.

[0089] The design of the terahertz / ultraviolet high-speed focal plane array is as follows:

[0090] The ultraviolet band FPA uses wide-bandgap materials such as AlGaN and ZnO to support short-wave UV-C sensing.

[0091] Response wavelength: 200–280 nm;

[0092] Charge collection time t c <100ns;

[0093] Quantum efficiency QE≈30–45%;

[0094] Resolution: 1024×1024 pixels;

[0095] Operating temperature: room temperature or TEC controlled cooling (-20℃);

[0096] Terahertz-band FPAs employ electrothermal detectors (Golay Cells), photoconductive arrays (PCAs), or two-dimensional graphene absorption layers.

[0097] Response bandwidth: 0.1–5THz;

[0098] Response time constant τ < 1 μs;

[0099] The thermoelectric voltage signal V = α·ΔT, where α is the thermoelectric coefficient.

[0100] According to an embodiment of the present invention, an illumination source is obtained by outputting continuous pulse illumination in the terahertz or ultraviolet band from a non-visible light source, specifically including:

[0101] The illumination source is obtained by outputting continuous pulse illumination in the terahertz or ultraviolet band from a non-visible light source, specifically including:

[0102] Based on the continuous wave output in the terahertz band by quantum cascade laser, the continuous wave is converted into a pulse sequence based on electro-optic modulation;

[0103] Terahertz pulsed illumination can be obtained by adjusting the pulse width based on the pulse sequence;

[0104] Based on the ultraviolet laser emitting ultraviolet light, the pulse parameters of the ultraviolet light are set, including the pulse frequency and pulse band;

[0105] Continuous pulse illumination in the ultraviolet light band is obtained based on the pulse frequency and pulse band of ultraviolet light;

[0106] The illumination source is obtained by weighted fusion of terahertz pulsed illumination and ultraviolet continuous pulsed illumination.

[0107] like Figure 2 As shown, the incident light is modulated through a periodic subwavelength structure, and the output wavefront is focused or dispersed behind the array. The geometric parameters of each structural unit can be independently controlled to manage its phase delay response, enabling arbitrary in-plane phase distribution design while maintaining both high transmittance and broadband response. The array is suitable for imaging systems with wavelengths from 200 nm to 10 THz. According to an embodiment of the present invention, the metasurface broadband lens design method is as follows:

[0108] Metasurface lenses achieve holographic manipulation of wavefronts by arranging nanostructures on a subwavelength scale. The phase response is analyzed using the principle of equal phase delay, as shown in the following formula:

[0109]

[0110] in:

[0111] φ(x,y,λ): The phase response of the metasurface unit at wavelength λ;

[0112] (x0, y0): Coordinates of the focus;

[0113] f: focal length;

[0114] λ: Incident wavelength (200nm–3mm);

[0115] The metasurface material employs a two-dimensional periodic array, and the phase delay is automatically adjusted according to λ to form a lens array with controllable dispersion.

[0116] According to an embodiment of the present invention, an incident light source is irradiated onto the target surface, and a sequence of target reflection images is acquired based on a high-speed focal plane array detector, specifically including:

[0117] Based on the incident visible light signal obtained from the incident light source, let the incident non-visible light signal be:

[0118] I in (x,y,λ,t)=I0(x,y,λ)·e -μ(λ)d

[0119] I0: Initial incident irradiance;

[0120] μ(λ): Wavelength-dependent absorption coefficient;

[0121] d: Thickness of the material being measured;

[0122] After being manipulated by a metasurface lens, the image is focused onto the plane of a high-speed focal plane array detector, resulting in a sequence of target reflection images, as shown in the following formula:

[0123]

[0124] x': Horizontal spatial coordinates on the imaging plane (usually the horizontal position of a pixel in a focal plane detector);

[0125] y': Vertical spatial coordinates on the imaging plane (usually the vertical position of a pixel in a focal plane detector);

[0126] t: Time variable (corresponding to a frame at a certain moment in the high-speed imaging sequence), used to describe the dynamic changes of the target over time.

[0127] T(λ): Transmittance of the metasurface;

[0128] δ(φ): Phase matching function;

[0129] I foc : Brightness of the imaging point.

[0130] According to an embodiment of the present invention, a target reflection image sequence is preprocessed based on a sparse image processing algorithm to obtain a complete frame sequence of target reflection images, specifically including:

[0131] The target reflection image sequence is obtained, and the target reflection image sequence is decomposed into low-frequency approximate components and high-frequency detail components based on wavelet transform.

[0132] Based on the high-frequency detail components, sparse distribution information is obtained. Based on the Fourier transform, the low-frequency approximation classification of the low-frequency region is transformed into the frequency domain to obtain the sparse coefficients of the high-frequency region.

[0133] By quantizing and encoding the sparse distribution information and the sparse coefficients in the high-frequency region, compressed sparsity is obtained.

[0134] Based on compression sparse analysis, compression ratio and distortion state information are balanced and distributed to obtain the optimal compression ratio, and target reflection images of complete frame sequences are generated.

[0135] It should be noted that the image processing uses the following joint optimization objective:

[0136]

[0137] Where: x t-1 This represents the image of frame t-1, used for the temporal difference sparse regularization term (which utilizes the small changes between adjacent frames to constrain the reconstruction).

[0138] λ1 represents the weight coefficient of the corresponding data consistency constraint, which controls the degree of fit between the reconstruction result and the compressed sampled data.

[0139] λ2 represents the weight coefficient of the corresponding sparse regularization term, which controls the sparsity constraint strength of the image in the sparse domain (such as wavelet, DCT).

[0140] λ3 represents the weight coefficient of the corresponding image total variation (TV) regularization term, which is used to preserve edge structure and suppress noise.

[0141] x t : The image of frame t;

[0142] y t : Corresponding compressed sampling data;

[0143] Φ: Sampling matrix;

[0144] Ψ: Sparse transformation basis;

[0145] Total variation of the image.

[0146] High frame rate imaging data compression modeling:

[0147] At frame rates exceeding 10,000 fps, the amount of data generated per second is extremely large, for example:

[0148] Data rate = R x ·R y ·D b ·f s

[0149] in:

[0150] R x ,R y Image resolution (e.g., 1024×1024);

[0151] D b : Number of bits of data per pixel (e.g., 12 bits);

[0152] f s Frame rate (e.g., 10,000 fps);

[0153] For example:

[0154] 1024×1024×12bits×10 4 fps = 1258Gbps

[0155] The amount of data far exceeds existing readout and transmission links (such as MIPI / PCIe). Therefore, sparse image processing and selective storage mechanisms must be used for compression.

[0156] Spatial domain compressed sensing modeling involves compressing data based on a compression model, as follows:

[0157] Each frame image Mapping to a sparse domain:

[0158] x = Ψ -1 ·s,‖s‖0<<N

[0159] Ψ: Wavelet / DCT transform matrix;

[0160] s: Sparse representation coefficients.

[0161] Using random measurement matrix sampling:

[0162] y = Φ·x = Φ·Ψ -1 ·s

[0163] The reconstruction process involves solving:

[0164]

[0165] A is the sampling matrix, which represents the linear measurement matrix used to sample the original sparse signal in compressed sensing imaging.

[0166] According to an embodiment of the present invention, the target reflection image of a complete frame sequence is reconstructed to obtain a reconstructed image, and the reconstructed image is transmitted to a terminal for display in a preset manner, specifically including:

[0167] Image decompression of target reflection images of complete frame sequences is performed based on sparse regularization constraints to obtain decompressed data;

[0168] Set a segmentation threshold, and perform noise analysis on the pressurized data based on the segmentation threshold to obtain noise data;

[0169] Noise is removed from the data to obtain noise-free data. Dynamic reconstruction is then performed based on the noise-free data to obtain the reconstructed image.

[0170] It should be noted that sparse decoding and temporal difference regularization are used to recover the complete frame sequence. The temporal difference sparse regularization is as follows:

[0171] Since non-visible light imaging objects (such as wafer surfaces and biological tissues) exhibit small inter-frame variations, temporal difference regularization is more suitable.

[0172]

[0173] This represents the temporal difference regularization term, which is the difference sparsity constraint between two adjacent frames. It is used to emphasize the characteristic of small inter-frame changes, thereby suppressing noise and preserving temporal consistency during reconstruction.

[0174] T represents the time dimension (i.e., the time index range of the frame sequence), which is used in the formula to accumulate the differences between adjacent frames to form the overall time difference regularization term.

[0175] In conjunction with image structure preservation, a total variation (TV) term is introduced:

[0176]

[0177] x i,j The pixel value (which can be a grayscale value or the value of a certain channel) in the image at the i-th row and j-th column.

[0178] x i+1,j : The pixel value of the image in the (i+1)th row and the jth column, which is the vertically adjacent pixel of xi,j.

[0179] x i,j+1 : The pixel value of the image in the i-th row and j+1-th column, which is the horizontally adjacent pixel of xi,j.

[0180] Final optimization goal:

[0181]

[0182] This application supports adaptive frame rate adjustment, with the sampling rate based on the inter-frame difference factor:

[0183]

[0184] When D t When the current frame is less than τ, the system automatically skips the storage of the current frame and compresses the output frame stream.

[0185] According to an embodiment of the present invention, the image reconstruction process further includes: keyframe prediction and inter-frame interpolation reconstruction.

[0186] If bandwidth is limited, inter-frame interpolation reconstruction can further compress the data volume:

[0187] Only keyframes x1, x4, x7, ... are sampled;

[0188] Recover intermediate frames using spatiotemporal interpolation (such as depthwise separable convolution or B-spline interpolation):

[0189]

[0190] in:

[0191] α∈[0,1]: time interpolation factor;

[0192] δ t Edge correction term (estimated via CNN).

[0193] Frame rate adaptive mechanism

[0194] The system dynamically adjusts the frame rate and compression rate based on the scene:

[0195] Using image difference rate D t =‖x t -x t-1 || / ||x t-1 ||;

[0196] When D t When the value is less than τ, skip frame storage and only update the trajectory;

[0197] Dynamic frame rate switching is possible. Without losing key information.

[0198] like Figure 3 As shown, in a second aspect, embodiments of this application provide a high-speed imaging system for the non-visible light band based on metasurface lenses. This system includes a memory and a processor. The memory includes a program for a high-speed imaging method for the non-visible light band based on metasurface lenses. When the program for the high-speed imaging method for the non-visible light band based on metasurface lenses is executed by the processor, it implements the following steps:

[0199] An illumination source is obtained by outputting continuous pulse illumination in the terahertz or ultraviolet band from a non-visible light source;

[0200] An incident light source is obtained by controlling the illumination light source using metasurface lenses;

[0201] An incident light source is irradiated onto the target surface, and a sequence of target reflection images is acquired based on a high-speed focal plane array detector.

[0202] The target reflection image sequence is preprocessed based on the sparse image processing algorithm to obtain the target reflection image of the complete frame sequence;

[0203] The target reflection image of the complete frame sequence is reconstructed to obtain the reconstructed image, which is then transmitted to the terminal for display according to a preset method.

[0204] It should be noted that this system comprises the following components:

[0205] Non-visible light ultrafast light source module, used to output terahertz or ultraviolet band continuous / pulse illumination;

[0206] Metasurface broadband lens arrays are used to achieve focusing, phase control, and dispersion modulation of incident waves;

[0207] High-speed focal plane array detector (HS-FPA) is used to receive images of target reflections after lens modulation;

[0208] The synchronous acquisition and control module is used to control exposure timing and data buffering under frame rate ≥ 10,000fps conditions; the synchronous acquisition and control module must meet the following requirements:

[0209] Trigger frequency: f s ≥10,000fps;

[0210] Exposure time control: t exp <50μs;

[0211] The light source synchronization trigger signal S(t) satisfies:

[0212]

[0213] The image synchronization readout logic is implemented using an FPGA, and the data is cached in DDR or SoC.

[0214] The image reconstruction and processing module is used to complete image decompression, denoising, and dynamic reconstruction based on sparse regularization constraints.

[0215] Specifically, the non-visible light source module outputs THz / UV band illumination light, which is focused and controlled by a metasurface lens array to illuminate the sample under test. The reflected / transmitted waves are collected by a high-speed focal plane detector and synchronously transmitted to the FPGA control module for data caching and frame synchronization. The image is then sparsely restored and temporally reconstructed by the reconstruction module, and finally visualized or analyzed through the output interface.

[0216] like Figure 4 As shown, this invention discloses a cross-sectional view of an ultraviolet / terahertz FPA chip structure. The cross-sectional view shows the chip material composition, the distribution of the photosensitive layer and the readout circuit, in order to illustrate the nanosecond-level response capability.

[0217] like Figure 5 As shown, the visual effect of this system in detecting microcracks and impurities under 12,000fps ultraviolet imaging conditions is compared with that of a traditional system. The system supports frame rates exceeding 10,000fps and uses sparse image processing algorithms to compress the data stream, achieving high frame rate, wide bandwidth, and non-destructive imaging, suitable for scenarios such as semiconductor wafer defect detection and non-destructive visualization of biological tissues.

[0218] According to an embodiment of the present invention, an illumination source is obtained by outputting continuous pulse illumination in the terahertz or ultraviolet band from a non-visible light source, specifically including:

[0219] The illumination source is obtained by outputting continuous pulse illumination in the terahertz or ultraviolet band from a non-visible light source, specifically including:

[0220] Based on the continuous wave output in the terahertz band by quantum cascade laser, the continuous wave is converted into a pulse sequence based on electro-optic modulation;

[0221] Terahertz pulsed illumination can be obtained by adjusting the pulse width based on the pulse sequence;

[0222] Based on the ultraviolet laser emitting ultraviolet light, the pulse parameters of the ultraviolet light are set, including the pulse frequency and pulse band;

[0223] Continuous pulse illumination in the ultraviolet light band is obtained based on the pulse frequency and pulse band of ultraviolet light;

[0224] The illumination source is obtained by weighted fusion of terahertz pulsed illumination and ultraviolet continuous pulsed illumination.

[0225] According to an embodiment of the present invention, the design method of a metasurface broadband lens is as follows:

[0226] Metasurface lenses achieve holographic manipulation of wavefronts by arranging nanostructures on a subwavelength scale. The phase response is analyzed using the principle of equal phase delay, as shown in the following formula:

[0227]

[0228] in:

[0229] φ(x,y,λ): The phase response of the metasurface unit at wavelength λ;

[0230] (x0, y0): Coordinates of the focus;

[0231] f: focal length;

[0232] λ: Incident wavelength (200nm–3mm);

[0233] The metasurface material employs a two-dimensional periodic array, and the phase delay is automatically adjusted according to λ to form a lens array with controllable dispersion.

[0234] According to an embodiment of the present invention, an incident light source is irradiated onto the target surface, and a sequence of target reflection images is acquired based on a high-speed focal plane array detector, specifically including:

[0235] Based on the incident visible light signal obtained from the incident light source, let the incident non-visible light signal be:

[0236] I in (x,y,λ,t)=I0(x,y,λ)·e -μ(λ)d

[0237] I0: Initial incident irradiance;

[0238] μ(λ): Wavelength-dependent absorption coefficient;

[0239] d: Thickness of the material being measured;

[0240] After being manipulated by a metasurface lens, the image is focused onto the plane of a high-speed focal plane array detector, resulting in a sequence of target reflection images, as shown in the following formula:

[0241]

[0242] T(λ): Transmittance of the metasurface;

[0243] δ(φ): Phase matching function;

[0244] I foc : Brightness of the imaging point.

[0245] According to an embodiment of the present invention, a target reflection image sequence is preprocessed based on a sparse image processing algorithm to obtain a complete frame sequence of target reflection images, specifically including:

[0246] The target reflection image sequence is obtained, and the target reflection image sequence is decomposed into low-frequency approximate components and high-frequency detail components based on wavelet transform.

[0247] Based on the high-frequency detail components, sparse distribution information is obtained. Based on the Fourier transform, the low-frequency approximation classification of the low-frequency region is transformed into the frequency domain to obtain the sparse coefficients of the high-frequency region.

[0248] By quantizing and encoding the sparse distribution information and the sparse coefficients in the high-frequency region, compressed sparsity is obtained.

[0249] Based on compression sparse analysis, compression ratio and distortion state information are balanced and distributed to obtain the optimal compression ratio, and target reflection images of complete frame sequences are generated.

[0250] It should be noted that the image processing uses the following joint optimization objective:

[0251]

[0252] in:

[0253] x t : The image of frame t;

[0254] y t : Corresponding compressed sampling data;

[0255] Φ: Sampling matrix;

[0256] Ψ: Sparse transformation basis;

[0257] Total variation of the image.

[0258] High frame rate imaging data compression modeling:

[0259] At frame rates exceeding 10,000 fps, the amount of data generated per second is extremely large, for example:

[0260] Data rate = R x ·R y ·Db ·f s

[0261] in:

[0262] R x ,R y Image resolution (e.g., 1024×1024);

[0263] D b : Number of bits of data per pixel (e.g., 12 bits);

[0264] f s Frame rate (e.g., 10,000 fps);

[0265] For example:

[0266] 1024×1024×12bits×10 4 fps = 1258Gbps

[0267] The amount of data far exceeds existing readout and transmission links (such as MIPI / PCIe). Therefore, sparse image processing and selective storage mechanisms must be used for compression.

[0268] Spatial domain compressed sensing modeling involves compressing data based on a compression model, as follows:

[0269] Each frame image Mapping to a sparse domain:

[0270] x = Ψ -1 ·s,‖s‖0<<N

[0271] Ψ: Wavelet / DCT transform matrix;

[0272] s: Sparse representation coefficients.

[0273] Using random measurement matrix sampling:

[0274] y = Φ·x = Φ·Ψ -1 ·s

[0275] The reconstruction process involves solving:

[0276]

[0277] According to an embodiment of the present invention, the target reflection image of a complete frame sequence is reconstructed to obtain a reconstructed image, and the reconstructed image is transmitted to a terminal for display in a preset manner, specifically including:

[0278] Image decompression of target reflection images of complete frame sequences is performed based on sparse regularization constraints to obtain decompressed data;

[0279] Set a segmentation threshold, and perform noise analysis on the pressurized data based on the segmentation threshold to obtain noise data;

[0280] Noise is removed from the data to obtain noise-free data. Dynamic reconstruction is then performed based on the noise-free data to obtain the reconstructed image.

[0281] It should be noted that sparse decoding and temporal difference regularization are used to recover the complete frame sequence. The temporal difference sparse regularization is as follows:

[0282] Since non-visible light imaging objects (such as wafer surfaces and biological tissues) exhibit small inter-frame variations, temporal difference regularization is more suitable.

[0283]

[0284] In conjunction with image structure preservation, a total variation (TV) term is introduced:

[0285]

[0286] Final optimization goal:

[0287]

[0288] This application supports adaptive frame rate adjustment, with the sampling rate based on the inter-frame difference factor:

[0289]

[0290] When D t When the current frame is less than τ, the system automatically skips the storage of the current frame and compresses the output frame stream.

[0291] According to an embodiment of the present invention, the image reconstruction process further includes: keyframe prediction and inter-frame interpolation reconstruction.

[0292] If bandwidth is limited, inter-frame interpolation reconstruction can further compress the data volume:

[0293] Only keyframes x1, x4, x7, ... are sampled;

[0294] Recover intermediate frames using spatiotemporal interpolation (such as depthwise separable convolution or B-spline interpolation):

[0295]

[0296] in:

[0297] α∈[0,1]: time interpolation factor;

[0298] δ t Edge correction term (estimated via CNN).

[0299] Frame rate adaptive mechanism

[0300] The system dynamically adjusts the frame rate and compression rate based on the scene:

[0301] Using image difference rate D t =‖x t -x t-1 || / ||x t-1 ||;

[0302] When D t When the value is less than τ, skip frame storage and only update the trajectory;

[0303] Dynamic frame rate switching is possible. Without losing key information.

[0304] A third aspect of the present invention provides a computer-readable storage medium including a program for a high-speed imaging method in the non-visible light band based on a metasurface lens. When the program for the high-speed imaging method in the non-visible light band based on a metasurface lens is executed by a processor, it implements the steps of the high-speed imaging method in the non-visible light band based on a metasurface lens as described in any of the above claims.

[0305] This invention discloses a high-speed imaging method, system, and medium for the non-visible light band based on metasurface lenses. The method involves obtaining an illumination source by outputting continuous pulse illumination in the terahertz or ultraviolet band from a non-visible light source; obtaining an incident light source by controlling the illumination source using metasurface lenses; illuminating a target surface with the incident light source; acquiring a sequence of target reflection images using a high-speed focal plane array detector; preprocessing the target reflection image sequence using a sparse image processing algorithm to obtain a complete frame sequence of target reflection images; reconstructing the complete frame sequence of target reflection images to obtain reconstructed images; and transmitting the reconstructed images to a terminal for display according to a preset method. The image is focused and controlled by a metasurface lens array before being illuminated onto the target surface and acquired by a high-speed focal plane detector. The response time is on the order of nanoseconds, adapting to high-speed processes. The integrated micro metasurface lens array improves wavefront control and focusing efficiency.

[0306] In the several embodiments provided in this application, it should be understood that the disclosed devices and methods can be implemented in other ways. The device embodiments described above are merely illustrative. For example, the division of units is only a logical functional division, and in actual implementation, there may be other division methods, such as: multiple units or components can be combined, or integrated into another system, or some features can be ignored or not executed. In addition, the coupling, direct coupling, or communication connection between the various components shown or discussed can be through some interfaces, and the indirect coupling or communication connection between devices or units can be electrical, mechanical, or other forms.

[0307] The units described above as separate components may or may not be physically separate. The components shown as units may or may not be physical units. They may be located in one place or distributed across multiple network units. Some or all of the units may be selected to achieve the purpose of this embodiment according to actual needs.

[0308] In addition, in the various embodiments of the present invention, each functional unit can be integrated into one processing unit, or each unit can be a separate unit, or two or more units can be integrated into one unit; the integrated unit can be implemented in hardware or in the form of hardware plus software functional units.

[0309] Those skilled in the art will understand that all or part of the steps of the above method embodiments can be implemented by hardware related to program instructions. The aforementioned program can be stored in a readable storage medium. When the program is executed, it performs the steps of the above method embodiments. The aforementioned storage medium includes various media capable of storing program code, such as mobile storage devices, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0310] Alternatively, if the integrated units of the present invention are implemented as software functional modules and sold or used as independent products, they can also be stored in a readable storage medium. Based on this understanding, the technical solutions of the embodiments of the present invention, or the parts that contribute to the prior art, can be embodied in the form of a software product. This software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as mobile storage devices, ROM, RAM, magnetic disks, or optical disks.

Claims

1. A high-speed imaging method for the non-visible light band based on metasurface lenses, characterized in that, include: An illumination source is obtained by outputting continuous pulse illumination in the terahertz or ultraviolet band from a non-visible light source; An incident light source is obtained by controlling the illumination light source using metasurface lenses; An incident light source is irradiated onto the target surface, and a sequence of target reflection images is acquired based on a high-speed focal plane array detector. The target reflection image sequence is preprocessed based on the sparse image processing algorithm to obtain the target reflection image of the complete frame sequence; The target reflection image of the complete frame sequence is reconstructed to obtain the reconstructed image, which is then transmitted to the terminal for display according to a preset method.

2. The high-speed imaging method for the non-visible light band based on metasurface lenses according to claim 1, characterized in that, The illumination source is obtained by outputting continuous pulse illumination in the terahertz or ultraviolet band from a non-visible light source, specifically including: The illumination source is obtained by outputting continuous pulse illumination in the terahertz or ultraviolet band from a non-visible light source, specifically including: Based on the continuous wave output in the terahertz band by quantum cascade laser, the continuous wave is converted into a pulse sequence based on electro-optic modulation; Terahertz pulsed illumination can be obtained by adjusting the pulse width based on the pulse sequence; Based on the ultraviolet laser emitting ultraviolet light, the pulse parameters of the ultraviolet light are set, including the pulse frequency and pulse band; Continuous pulse illumination in the ultraviolet light band is obtained based on the pulse frequency and pulse band of ultraviolet light; The illumination source is obtained by weighted fusion of terahertz pulsed illumination and ultraviolet continuous pulsed illumination.

3. The high-speed imaging method for the non-visible light band based on metasurface lenses according to claim 2, characterized in that, The design method for metasurface broadband lenses is as follows: Metasurface lenses achieve holographic manipulation of wavefronts by arranging nanostructures on a subwavelength scale. The phase response is analyzed using the principle of equal phase delay, as shown in the following formula: in: φ(x,y,λ): The phase response of the metasurface unit at wavelength λ; (x0, y0): Coordinates of the focus; f: focal length; λ: Incident wavelength (200nm–3mm); The metasurface material employs a two-dimensional periodic array, and the phase delay is automatically adjusted according to λ to form a lens array with controllable dispersion.

4. The high-speed imaging method for the non-visible light band based on metasurface lenses according to claim 3, characterized in that, An incident light source illuminates the target surface, and a sequence of target reflection images is acquired using a high-speed focal plane array detector. Specifically, this includes: Based on the incident visible light signal obtained from the incident light source, let the incident non-visible light signal be: HER in (x,y,λ,t)I0(x,y,λ)·e -μ(λ)d I0: Initial incident irradiance; μ(λ): Wavelength-dependent absorption coefficient; d: Thickness of the material being measured; After being manipulated by a metasurface lens, the image is focused onto the plane of a high-speed focal plane array detector, resulting in a sequence of target reflection images, as shown in the following formula: T(λ): Transmittance of the metasurface; δ(φ): Phase matching function; I foc : Brightness of the imaging point.

5. The high-speed imaging method for the non-visible light band based on metasurface lenses according to claim 4, characterized in that, The target reflection image sequence is preprocessed based on a sparse image processing algorithm to obtain a complete frame sequence of target reflection images, specifically including: The target reflection image sequence is obtained, and the target reflection image sequence is decomposed into low-frequency approximate components and high-frequency detail components based on wavelet transform. Based on the high-frequency detail components, sparse distribution information is obtained. Based on the Fourier transform, the low-frequency approximation classification of the low-frequency region is transformed into the frequency domain to obtain the sparse coefficients of the high-frequency region. By quantizing and encoding the sparse distribution information and the sparse coefficients in the high-frequency region, compressed sparsity is obtained. Based on compression sparse analysis, compression ratio and distortion state information are balanced and distributed to obtain the optimal compression ratio, and target reflection images of complete frame sequences are generated.

6. The high-speed imaging method for the non-visible light band based on metasurface lenses according to claim 5, characterized in that, The target reflection image of the complete frame sequence is reconstructed to obtain the reconstructed image. The reconstructed image is then transmitted to the terminal for display according to a preset method, specifically including: Image decompression of target reflection images of complete frame sequences is performed based on sparse regularization constraints to obtain decompressed data; Set a segmentation threshold, and perform noise analysis on the pressurized data based on the segmentation threshold to obtain noise data; Noise is removed from the data to obtain noise-free data. Dynamic reconstruction is then performed based on the noise-free data to obtain the reconstructed image.

7. A high-speed imaging system for the non-visible light band based on metasurface lenses, characterized in that, The system includes a memory and a processor. The memory contains a program for a high-speed imaging method in the non-visible light band based on metasurface lenses. When the program for the high-speed imaging method in the non-visible light band based on metasurface lenses is executed by the processor, it performs the following steps: An illumination source is obtained by outputting continuous pulse illumination in the terahertz or ultraviolet band from a non-visible light source; An incident light source is obtained by controlling the illumination light source using metasurface lenses; An incident light source is irradiated onto the target surface, and a sequence of target reflection images is acquired based on a high-speed focal plane array detector. The target reflection image sequence is preprocessed based on the sparse image processing algorithm to obtain the target reflection image of the complete frame sequence; The target reflection image of the complete frame sequence is reconstructed to obtain the reconstructed image, which is then transmitted to the terminal for display according to a preset method.

8. The high-speed imaging system for the non-visible light band based on metasurface lenses according to claim 7, characterized in that, The illumination source is obtained by outputting continuous pulse illumination in the terahertz or ultraviolet band from a non-visible light source, specifically including: The illumination source is obtained by outputting continuous pulse illumination in the terahertz or ultraviolet band from a non-visible light source, specifically including: Based on the continuous wave output in the terahertz band by quantum cascade laser, the continuous wave is converted into a pulse sequence based on electro-optic modulation; Terahertz pulsed illumination can be obtained by adjusting the pulse width based on the pulse sequence; Based on the ultraviolet laser emitting ultraviolet light, the pulse parameters of the ultraviolet light are set, including the pulse frequency and pulse band; Continuous pulse illumination in the ultraviolet light band is obtained based on the pulse frequency and pulse band of ultraviolet light; The illumination source is obtained by weighted fusion of terahertz pulsed illumination and ultraviolet continuous pulsed illumination.

9. The high-speed imaging system for the non-visible light band based on metasurface lenses according to claim 8, characterized in that, The design method for metasurface broadband lenses is as follows: Metasurface lenses achieve holographic manipulation of wavefronts by arranging nanostructures on a subwavelength scale. The phase response is analyzed using the principle of equal phase delay, as shown in the following formula: in: φ(x,y,λ): The phase response of the metasurface unit at wavelength λ; (x0, y0): Coordinates of the focus; f: focal length; λ: Incident wavelength (200nm–3mm); The metasurface material employs a two-dimensional periodic array, and the phase delay is automatically adjusted according to λ to form a lens array with controllable dispersion.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium includes a high-speed imaging method program for the non-visible light band based on metasurface lenses. When the high-speed imaging method program for the non-visible light band based on metasurface lenses is executed by a processor, it implements the steps of the high-speed imaging method for the non-visible light band based on metasurface lenses as described in any one of claims 1 to 6.

Images